recommend ceiling mount access point simultaneous dual or tri band AC1900, 3200

[sfx2000]

At this moment, I don't believe any iOS devices support MU-MIMO.

As of August, 2016, there is no MU-MIMO support in any Apple equipment...

 

[L&LD]

As of August, 2016, there is no MU-MIMO support in any Apple equipment...

The 'magical' pixie dust is on back order still.

 

[sfx2000]

The 'magical' pixie dust is on back order still.

Maybe...

Or just that there's no pixies there...

Apple jumps into Wireless technology when it suits their vision of the User Experience - MU is not a big enhancement there - they were leaders in 802.11g/11n/11ac, and they deployed both clients and AP's to that end...

If MU was a big deal, you would see an Airport Extreme AC-MU, and clients launched with it, both on iOS and macOS... when Apple jumps into the pool, they jump in big-time - and this basically hasn't happened.

To be honest - MU hasn't been a huge deal, and a definite non-adder to individual client performance.

Residential/Consumer Router/AP OEM's are advertising MU as a big deal, as they want to rotate out the fielded inventory, e.g. sales of new equipment, so they put really big numbers on them - e.g. AC3100/AC5300/AC2400/etc... when most folks are totally good with a decent AC1900 AP...

AC1900 was a very big jump, and most folks did - but the OEM's are now stuck as most folks that wanted 802.11ac, they have it, and those solutions are "good enough" - most sales these days I would suggest are replacements, not new 11ac customers.

 

[L&LD]

Maybe...

Or just that there's no pixies there...

Apple jumps into Wireless technology when it suits their vision of the User Experience - MU is not a big enhancement there - they were leaders in 802.11g/11n/11ac, and they deployed both clients and AP's to that end...

If MU was a big deal, you would see an Airport Extreme AC-MU, and clients launched with it, both on iOS and macOS... when Apple jumps into the pool, they jump in big-time - and this basically hasn't happened.

To be honest - MU hasn't been a huge deal, and a definite non-adder to individual client performance.

Residential/Consumer Router/AP OEM's are advertising MU as a big deal, as they want to rotate out the fielded inventory, e.g. sales of new equipment, so they put really big numbers on them - e.g. AC3100/AC5300/AC2400/etc... when most folks are totally good with a decent AC1900 AP...

AC1900 was a very big jump, and most folks did - but the OEM's are now stuck as most folks that wanted 802.11ac, they have it, and those solutions are "good enough" - most sales these days I would suggest are replacements, not new 11ac customers.

Apple hasn't had magical experiences for a long time now (pixie dust or not).

MU can be a big deal in the right environment. Apple hasn't jumped 'big-time' on anything for too long.

 

[sfx2000]

MU can be a big deal in the right environment.

Please feel free to elaborate on that statement

 

[L&LD]

Please feel free to elaborate on that statement

I'm sure you already know. As an example? An auditorium or even a large room (think 'party') filled with handheld devices and appropriate hardware on both ends (the AP/routers and the clients).

 

[sfx2000]

I'm sure you already know. As an example? An auditorium or even a large room (think 'party') filled with handheld devices and appropriate hardware on both ends (the AP/routers and the clients).

MU won't solve that problem - next?

 

[L&LD]

MU won't solve that problem - next?

That's the exact problem it will 'solve'. Can you elaborate now?

 

[sfx2000]

That's the exact problem it will 'solve'. Can you elaborate now?

Most clients are not MU capable - next?

 

[lifespeed]

MU won't solve that problem - next?

How about one client streaming a hi-bitrate video, perhaps 20Mbps, from the local media server while a MU-MIMO VoIP smartphone is trying to make a phone call without dropping packets? Shouldn't they receive simultaneous data, rather than time-multiplexed?

 

[lifespeed]

Most clients are not MU capable - next?

If the whole argument is other devices aren't there yet, well, it has been the same forever. 64-bit processors without 64-bit software, the list goes on. Either the base station or the handset has to come first, bonus points if you can at least start with one of each.

 

[sfx2000]

Most clients are not MU capable - next?

and I'll expand on that - imagine if you did have an auditorium full of MU clients - hundreds of them - Each AP can only support four client groups at any given time, and then...

It's the dream of multicast - it works for some cases - in WiFi, perhaps not - LTE/LTE-Advanced is in a better place there, as is 802.16, where one has a scheduled MAC, but there, the development costs increase dramatically...

MU-MIMO in WiFi - nice from a technical perspective, but from a practical/pragmatic view - it's a narrow use case...

next?

 

[sfx2000]

How about one client streaming a hi-bitrate video, perhaps 20Mbps, from the local media server while a MU-MIMO VoIP smartphone is trying to make a phone call without dropping packets? Shouldn't they receive simultaneous data, rather than time-multiplexed?

You're assuming that a 20 dollar chipset is capable of handling that on the 802.11 layer, and that the Router can manage traffic to that level in a 200 dollar box.

Not true at the current moment...

next?

 

[sfx2000]

If the whole argument is other devices aren't there yet, well, it has been the same forever. 64-bit processors without 64-bit software, the list goes on. Either the base station or the handset has to come first, bonus points if you can at least start with one of each.

64bit/32bit - many align with the improvements that we saw with x86 moving that way...

x86 had issues that long mode pretty much fixed - and RISC didn't benefit in that way - they did benefit from bigger memory space, but at a cost of larger binaries...

More improvement was in architecture efficiency, which would have happened in any event - case in point - Cortex-A53 compared to Cortex-A7 - running 32 bit code is still faster on the A53 - not because it's 64bit, it's just runs 32-bit code faster due to lessons learned...

 

[lifespeed]

You're assuming that a 20 dollar chipset is capable of handling that on the 802.11 layer, and that the Router can manage traffic to that level in a 200 dollar box.

Not true at the current moment...

next?

I'm not so sure of that, I have had very good results on my network with VoIP and other high-bandwidth applications. I chose my router for it's hardware NAT bandwidth and VoIP performance, it applies QoS and works flawlessly. The *only* choke point in my network so far has proven to be the wireless AP.

You may be a little pessimistic of the benefits of this technology. A household with 3 or 4 simultaneous users could see high-bandwidth applications coexist pretty well on separate spatial channels, certainly better than time slicing.

 

[sfx2000]

To re-cap here - I've put a lot of thought and resources into investigation of MU - I've done the simulations, and FWIW, Tim's results on the primary/main site pretty much confirm what I've seen - not much benefit except for sales/marketing with the current state of the art..

 

[sfx2000]

I'm not so sure of that, I have had very good results on my network with VoIP and other high-bandwidth applications. I chose my router for it's hardware NAT bandwidth and VoIP performance, it applies QoS and works flawlessly. The *only* choke point in my network so far has proven to be the wireless AP.

Not sure how that ties into MU - I've got a very nice network that can do the same here... and probably better than what's in your house...

Get away from confirmation bias - most of what you state can be handled by a basic AC1900 Router/AP... in SU mode at that.

 

[lifespeed]

Get away from confirmation bias - most of what you state can be handled by a basic AC1900 Router/AP... in SU mode at that.

It goes without saying that any dual-band AC access point is better than my old single band N. I'm going to buy a piece of network gear today and keep it for 5 years during which time much of my gear will support the newer tech.

Now go back to driving your carbureted, drum-braked car with bias ply tires

 

[sfx2000]

Get away from confirmation bias - most of what you state can be handled by a basic AC1900 Router/AP... in SU mode at that.

And that's why I generally recommend AC1900 class Router/AP's for most folks - An RT-AC68U/R7000/WRT1900ac/airport extreme ac/whatever DLink does these days because I don't follow them very closely

Most folks don't need any more than that...

For those who are considering breaking things out - it's about balance and resources - and most dedicated AP's don't do MU - and for good reason.

Most of the bottlenecks in SOHO networks are not wireless actually - and those that are, it's taking a practical approach and moving the AP to where the traffic is...

 

[sfx2000]

Now go back to driving your carbureted, drum-braked car with bias ply tires

67 camaro, 64 Mustang, and a Honda S600 - those are on my to-do list - I own two of the three - still looking for a good mustang to complete that group...

And for those who don't know the Honda S600 - mine is a Roadster - and restored from a barn find...

https://en.wikipedia.org/wiki/Honda_S600