Recommended Wireless Suppliers

[DavR]

As for companies - I can see Digital Air from my house and they have always been brilliant for me. I've probably spent 20k with them over 3 schools and I've spent at least 10k with them personally doing side-project installs.

 

I did like the guy from Digitial Air when I saw them at BETT, seemed very knowledgeable and to have a good approach. Also, can't knock a free survey when you need three quotes!

[DavR]

Secondary with 5 separate buildings, 50 APs.

 

That's not coverage sufficient for 1:1 though, the school has stated no intention to go down that route, but if they did decide to we would probably need another 15-20.

 

Yikes. Yes, I can see why you'd want a solid enterprise solution for that kind of coverage.

 

I just need something that will run a class of 15 devices well, with a bit of future proofing.

[Boredguy]

As for companies - I can see Digital Air from my house and they have always been brilliant for me. I've probably spent 20k with them over 3 schools and I've spent at least 10k with them personally doing side-project installs.

We used Digital Air when we upgraded our WiFi just over 2 years ago (and are using them again to add some additional capacity now we have a new building going up) and they certainly were competitive in the price and have given us good support for a couple of minor issues since installation.

 

We have (or will have) 50 AP's covering 14 buildings and fields by October.

[Tesla]

Any recommended Unifi resellers about? We're also potentially looking at new wireless.... at some form of budget.

[Oaktech]

Any recommended Unifi resellers about? We're also potentially looking at new wireless.... at some form of budget.

 

Digital Air are Unifi resellers, so yes, Digital Air!

[Tesla]

Digital Air are Unifi resellers, so yes, Digital Air!

 

Couldn't see anything for them on there. Ta!

[Oaktech]

Couldn't see anything for them on there. Ta!

 

Wifi Gear and Digital Air are the same people - Digital Air being the service and consultancy sidde and Wifi Gear basically being box shifters!

 

Ubiquiti

[dapaulio]

We buy from dabs, servium and 3ctech. Bought a load of ac pro for £110 each

[call911]

Unifi all the way. I a have over 300 deployed between three school buildings, and zero issues. Supports 1:1 devices, teachers devices, BYOD, etc. works great. We push about 800gb of traffic on the wireless every day.

[jmak]

We have Unifi and it's been solid for 2 years now. I've got the controller running on a VM, but I barely touch it. I just have a look every day to see who's turned theirs off! (I was on a very low budget, so the PoE injectors are plugged in in the classroom). You can run the controller on an AWS server if you want easier remote access/more resilience and the cost would be minimal as it's very small and not much data. I'd say it's not essential, as the APs will run without the controller, so you just lose visibility and the ability to change the config.

 

I wanted to see it working before I committed to buying - our local secondary run it and they have 1800 pupils and ~300 staff. They don't have 1:1 devices, but there's a decent number and coverage is great. The only hint they gave me was that they'd found that if the injectors were too far from the AP, there wasn't sufficient power, but their NM was very confident about them.

 

We went for the lowest spec devices, so only have b,g,n, but the basic units are now dual band dual radio.

 

I'm sure the pricier options are better, but whether you need the extras is debatable. They do the basics well and also support VLANs, multiple SSIDs, guest portals etc. At the price there's hardly anything to lose by giving them a try. I had them set up in minutes.

Edited June 11, 2016 by jmak

[rsaddul1]

Hello,

 

I was wondering if you could shed any light on an issue we currently have at our school.

 

We are an option 1 LGFL school, therefore our firewall is looked after by LGFL. We have an external vendor called Lan3 which currently support our Aerohive

(WiFi System).

 

We don't control our firewall as it is managed by LGFL. Unfortunately they don't support and won't configure BootP or DHCP-Relay.

 

This has now put us in a predicament since we have all traffic routed through the firewall. Bare in mind we are using VLANS for the WiFi and we don't want to result in having everything on a default vlan.

 

How did you manage to get this working?

 

Regards

Rhys

[DavR]

Hi Rhys,

 

Not sure I follow - what exactly isn't working for you?

 

From inference, I'm guessing you're not getting DHCP to your WiFi VLAN? When you say you've got all traffic routed through the firewall, do you mean you're using it as your Layer 3 switch to route betwen VLANs? How are your other VLANs getting DHCP?

 

We don't have ours VLAN'd off, but if I did I'd have my own L3 switch to do it rather than rely on the LGfL firewall, which under current limitations doesn't sound capable. Either that, or you'll have to stick all of your APs on static IP addresses.

 

The only problems we had with our LGfL firewall is we did need to open up a bunch of ports to allow APs to communicate with HiveManager online, but we got there in the end.

[AlanD]

I am going to weigh in here - and reiterate my beliefs - which is: You need to get some kit in and test it. I'm not saying that "sellers" will sell you something inappropriate - but you need to be in an informed position to negotiate. Yes - Ubiquiti is "cheap" - but don't let the price put you off, or let the ignorant rubbish it. It clearly works well in lots of schools - and has no on going licence costs and is fully functional with VLANs, RADIUS, AD, load balancing, band steering, airtime fairness and all the rest of it. But don't take my word that it would do the job either. Be clear about what you want to achieve. There are multitudes of myths and false beliefs and expectations - and sometimes downright lies surrounding what wireless can do. Don't get blinded by multigigabit connection speeds quoted in the literature. You are not going to achieve those - even with the latest most expensive kit. 11Mb/s "b" connections produced real world data rates of 2 to 3Mb/s. Fine for a single PC webbrowsing - but useless for a class. With 13 channels you could, however, use multiple APs easily. 54Mb/s "a/g" connections will deliver somewhere between 20 and 30Mb/s of real data - but use 20Mz channels - rather inefficiently and you could usually manage to use channels 1,4,9,and 13. (Remember that wireless is a half duplex medium, and has no means to avoid packet collisions other than random back off times) 150Mb/s "n" connections were an ingenious result of applying mathematics to the problems of wireless transport and allowing the use of wider 40MHz channels - and more than doubles the real data rate to around 80 or 90Mb/s. One slight downside is that because it makes really good use of the frequency channels you can now only use channels 1 , 6 and 11 on the 2.4GHz band. (There might be some merit in using 1, 7 and 13 - but not all wireless clients like anything above channel 11 - because 12 and 13 are not used outside the USA). "n" technology was the first that allowed us to get anywhere near using laptops in classrooms. But its far from ideal. Sharing even a 100Mb/s between 30 laptops doing a domain login and loading a user profile can be painfully slow. In fact real world data rates suffer quite a lot when you have 30 clients hitting the airways at once because collisions become increasingly detrimental. (I'll mention here that MERU has a technology that minimizes this degradation ...although while it offered advantages with densely packed classrooms using 2.4Ghz - I couldn't see it made much difference with today's wireless standards). After that you can add two or three streams - assuming that your client supports them. Don't go all starry eyed and believe that this is going to give you a data rate of 300 or 450Mb/s. Its only going to be a theoretical bit rate connection. It relies on spatial separation - and gain in real world data rates is only going to be around 20% at best. So now we are up to 110Mb/s - maybe 120Mb/s. A gain worth having - but its not transformational. The we had "ac1" (first wave of "ac"). While this has beamforming and higher QAM, its main potential advantage is wider channels of 80Mhz (or possibly 160Mhz). The trouble is that its almost impossible to deploy these wide channels sensibly in classroom blocks as there are relatively few 80Mhz channels. Turning the power right down (remember to do this on clients as well as access points) can help - and yes, access points and clients will to some extent attempt to arbitrate channels and widths in an attempt to optimise their capacity....but if you have heavy use in all these classrooms you might well find you are best sticking with 40MHz channels. The result is that real world data rates are still under 150Mb/s. Worse still, is that the fact that most ac1 clients are single stream only - partly because it saves on battery charge. So expect to lose 20% - which might be all the gain you achieved in moving from "n" to "ac1". Don't misunderstand me here: for a smaller number of connecting clients and using wider channels you will get data rates of up to 300Mb/s - but not where you need them most - in a cluster of heavily used classrooms. "ac2" has up to 8 streams - and adding streams obeys laws of diminishing returns so another 20% at best - but only in a world that doesn't exist because there are no 8 stream clients. It also offers MU-MIMO. This is a real progress. But don't get taken in by pictures showing you 4 clients each being served by a single stream. Just like any spatial stream system - you need to remember the waves of the streams are on the same frequency - and much of the potential gains are lost. However - you will get around a 20% overall increase in data rates - and that's worth something. There are schools that have taken out fixed ICT suites and use classrooms of laptops instead. They are on a steep learning curve. Yes - you "can" get high wireless speeds, and there are some gains to be had with latest "ac2" technology....but its not transformational but I think you need to think twice before throwing money at equipment believing it will solve the problems of wireless. You might well find that better positioning of access points (typically 1 per classroom - but it depends on your walls), perhaps replacing the wireless cards in your client devices, etc will bring you equal if not better benefits for your users.

[Mark-wilde]

Have bought Unifi POE switch, APs and several sfp modules.... very nicely used

[AlanD]

 

...you must be at one of those schools benefiting from the cuts elsewhere then....

 

When you say "crazy fast" - so you have any figures to back that up? Even with 80MHz channels - and you would be struggling to use those channel widths if there are neighbouring classrooms using wireless - the best you are going to get is around 200Mb/s of actually data (...forget those "connection" speeds on the dashboard).....Not sure I would describe that as "crazy fast"...But yes, "ac" with wide channels can do twice that of an "n" access point.

[zag]

...you must be at one of those schools benefiting from the cuts elsewhere then....

 

When you say "crazy fast" - so you have any figures to back that up? Even with 80MHz channels - and you would be struggling to use those channel widths if there are neighbouring classrooms using wireless - the best you are going to get is around 200Mb/s of actually data (...forget those "connection" speeds on the dashboard).....Not sure I would describe that as "crazy fast"...But yes, "ac" with wide channels can do twice that of an "n" access point.

 

We are not a rich school at all, but wireless connectivity is probably the most important infrastructure we have purchased. Bear in mind the wireless has been going for over 10 years now, with AB/G then N then AC Ap's over that time so it does average out over the lifetime of the system. Aruba has no ongoing costs either which is nice.

 

In terms of speed, I measure it in complaints really. It just works and teachers don't even notice any speed problems whch is a sign of its success.

 

Here's a snapshot of whats happening right now. The majority of users connect at 100-300Mb/s but as you can see, some can go up to 800Mb/s with the right hardware. I believe most of those are student Macbook Pro's or Airs which have the 3 antennas inside them.

 

 

Anyway, quite happy with the Aruba system. Its the clients that are the bottleneck more than anything.

Edited May 4, 2017 by zag

[Martin48]

2 sites here:

Aerohive installed by LAN3 highly recommended, no issues even use it to diagnose our aging lan network.

Unifi installed by XMA is ok does the job for the budget.

[AlanD]

"The majority of users connect at 100-300Mb/s but as you can see, some can go up to 800Mb/s with the right hardware"

 

As long as we all understand that "connecting" at is nothing like actual data rates. 3 streams only gives you around 20% more than a single stream in ideal scenarios (Ceiling mounted AP in the room) ...and its much the same for MU-MIMO. Yes the "connection" speeds are 3 times higher - but its like watching 3 televisions at once - and then trying to repeat what was said on each TV....Yes - you will mange more than was said on one TV - but you will struggle to get significantly more. I won't say multiple streams are a gimmick- because they do offer an increase in throughput - and possibly more usefully - work in hostile environments rather better and for longer distances...but quoting "gigabit speeds" is largely unhelpful and misrepresents what wireless is capable of. I've not seen actual data rates of more than around 200Mb/s (Even with a 1600Mb/s connection speeds)...and that is using wide 80MHz channels - which you can't use in classroom blocks without a significant co-channel hit....in which case you are better using 40MHz channels which an "n" AP can provide.

 

In a "lightly" loaded classroom - "AC" Access points will conspire to use wide 160Mhz and 80MHz channels effectively - but this drops back as soon as you have a class all using devices/laptops to a total bandwidth of around 100Mb/s (same as for an "n" Access point").

 

And you are most unlikely to benefit from aggregating 2 CAT5/6 cables to a single access point...which I have heard some have been advised to do.

 

 

..and I think the latest ipads only support 2 streams - and can't manage the 160MHz channel either....which is around 800Mb/s "Connection" speed....which is around 200Mb/s in real money....you/'d probably only get 10% extra (or less) with an third stream anyhow.

 

 

...and lost of quite high spec laptops are single stream...possibly because they think wider channels (which typically can't make much use of in schools) is now providing a faster data rate so they needn't bother with spatial streams (which helps save on battery too) ...which makes them poorer in noisy environments