I thought I would jot down a few thoughts on market trends.
2×2:2 Client Devices
The market is trending toward 2×2:2 clients. Those 4×4:4 clients that you were hoping for aren’t coming. In fact, 3×3:3 clients are are mostly on their way out, albeit slowly. Smartphones range from 1×1:1 to 2×2:2, and most tablets and laptops are already 2×2:2.
If you haven’t checked out my presentation on MU-MIMO, now’s your chance. Most of today’s 11ac Wave-2 APs are 4×4:4, but the landscape will soon trend toward 2×2:2 Wave-2 APs. We are years (as of this writing) from seeing any appreciable benefit from MU-MIMO, so why spend extra money on 4×4:4 APs? Good network design, using lower cost 2×2:2 Wave-2 APs, will yield an equally good user experience for far less money for the next AP lifecycle.
That said, vendors usually add additional value, beyond simple connectivity, to their high-end APs (however “high-end” that might be). It’s these additional benefits that should be weighed carefully when deciding on the APs to purchase.
More 5GHz radios, and thus more APs, is also the trend because high density requirements are popping up all over the place. For example, one of my K-12 customers has recently targeted a minimum of 90 devices per classroom for his design, and he fully expects to hit that target within the lifetime of his current generation of APs.
Buying more APs is a bit easier on the budget when you don’t need to buy super-high-end APs. The “street” cost difference between 4×4:4 and 3×3:3 APs is significant, even if all of them are Wave-2.
Dual-5GHz (D5G) APs
D5G APs promise to add more capacity for high density environments, at the same price. Unlike the current MU-MIMO hype, this would be real capacity gain, and vendors like Aerohive, Cisco, Xirrus, Motorola, Meru, and others are making strides in this area. The question at hand is whether or not each vendor’s D5G can actually deliver on the promise of higher capacity.
Fitting within a .3af PoE budget, without crippling the AP’s capabilities, requires going with 3×3:3 or less on dual-radio APs.
Single Radio Wave-2 APs
I wouldn’t mind seeing a single-radio, low-cost 3×3:3 Wave-2 AP that can accept PoE+ (.3at) and pass through PoE (.3af).
* It could be used as a low-cost way to add capacity without any additional IDF cable runs
* It could be used as a network connection (with PoE) point for cameras and other PoE powered IoT stuff
* You could string two of these APs in series for low-cost, high-throughout VHD deployments
Going forward, I expect to see all sorts of creative ways of implementing more 5GHz radios and more creative ways of negating the ACI that comes with close-quarters 5GHz radios.
A Plethora of Radio Types
Of course it’s a little scary, but it’s coming. Change is scary sometimes, but we gotta role with it. Vendors are already packing BLE beacons into APs, and cellular is next. I wouldn’t be at all surprised to see 11ad slip in beside 11ax in 2017. It’s a battle for “share of ceiling space”, and vendors will justify Franken-APs in all sorts of ways. Vendor marketing teams will have a field day with all of the new messaging possibilities.
Think it’s crazy? Look at your phone. They already have WiFi, BT, BLE, NFC, and quad-band cellular…and it won’t stop there. Those radios open Pandora’s box for WiFi infrastructure vendors because they’re already on the ceiling. Don’t believe me? Check out this non-sense: https://www.ubnt.com/unifi/unifi-ap-ac-edu/ Yes, that’s a speaker system integrated within the APs. It’s only a matter of time before it gets a little whacky. Integrated sprinkler systems perhaps? 🙂 I jest.
Share of ceiling space is a big reason (though not the only reason) for vendor consolidation, whereby all of the WiFi vendors are being scooped up. WiFi itself has always been a spear-head (door opener) for Ethernet sales with the big vendors…and that will only increase as the number and type of radios within an AP increases.
We’ve all been witnesses to the consolidation this far:
* Cisco buys Airespace, then Meraki
* Juniper buys Trapeze (I’ll skip Trapeze’s long backstory)
* Aruba buys NetChem, AirWave, Avenda, Azalea, and AmigoPod, then HP buys Aruba
* Avaya kinda/sorta buys Xirrus
* Brocade buys Ruckus, who bought Cloudpath and Yfind Technologies
* Zebra buys Motorola, who bought AirDefense and Wireless Valley
* NetScout bought Fluke, who bought AirMagnet
* AirTight Networks…changed its name to Mojo Networks #WhateverThatMeans
Aerohive will get bought. Soon. For a bigger-than-expected number.
Ubiquiti will try to step up into the SME.
Companies like Nyansa will enter the market to innovate in areas that infrastructure vendors haven’t, filling important feature gaps. They’ll get bought early as a “tuck in.”
In 5 years, it’ll be a different ballgame, but 90% of WiFi networks will still suck. How long did it take before Ethernet networks stopped sucking? Oh yeah, many of them still do…after nearly 40 years of “practice.”