Blogs

2.4GHz is Dead! Long Live 2.4GHz!

By Radzima posted Oct 29, 2014 12:01 PM

  

This past week on Twitter I witnessed a brief “discussion” between some wireless folks regarding the elimination of 2.4GHz radios in APs and the creation of a 5GHz Wi-Fi utopia. Now I’ve had this conversation many times before and I’ve had neither the popular opinion nor the ability to sway the minds of self-proclaimed saviors of the industry. That’s not to say it’s entirely impossible, it’s just wholly impractical for nearly every enterprise out there and would create more issues than it would solve.

 

Let’s start with the basic premise of the argument: 2.4GHz is bad for everyone from designers to users.

 

“But why is it bad?” you ask. Well, to put it not so gently the 2.4GHz ISM band is an overcrowded wasteland filled with all manner of wireless communications and RF bleed. From Bluetooth to Zigbee to cordless phones, microwaves to fluorescent light ballasts, all topped off with limited channels, aging technology, and required backwards compatibility slowing everything down, it’s full of things that keep wireless engineers up at night. The deployment of said frequencies in enterprise WLANs leads to frequent and intermittent issues which leads to poor client performance which in turn leads to poor user perception. The migration of all Wi-Fi to the green pastures and open fields of 5GHz will ease design, planning, and management, improve client performance and user perception, as well as lead to cheaper APs by removing the need for that nasty 802.11b/g/n radio.

 

Fair point. No argument from me there. In fact, I doubt you’ll find an argument from anyone against that. So that’s it, right? Down with 2.4GHz! Problem solved!

 

Not so fast… Remember the biggest driver for enterprise Wi-Fi adoption in recent history? The acronym dreaded by more enterprise engineers than PCI-DSS? That’s right, BYOD. “But devices are built dual-band now” you say. “We can’t support everything under the sun” you cry. I feel for you, I really do, but that doesn’t change the landscape of budget friendly client devices crowding the market. For example, on a whim I recently purchased a mid-range Windows 8.1 laptop on sale for a ridiculously low amount of money because why not? Actually, I owned exactly zero Windows machines and needed something to test with in my lab and the price/performance ratio was enough to entice me. I gave the specs a brief glance and decided it’d be good enough to mess around with so I happily clicked “Buy” without a second thought. Once the machine was delivered and I started to get it setup with all my tools I noticed pretty quickly it wasn’t seeing any of my 5GHz-only networks. This is brand new, out of the box, big name laptop (not some no-name manufacturer repackaging last year’s models) and in 2014 it came with a 1x1 802.11b/g/n adapter installed.

 

Out of curiosity I did a little poking around on the sites of two major manufacturers while writing this to see just how bad it is and found that nearly all the laptops, even the “performance” models, were offered with single-band 802.11b/g/n adapters. To make matters worse, those basic cards were only 1x1 too. On the models that even had the option to upgrade to a dual-band adapter it added anywhere from $15-$50. Tablets gets even worse. On one OS maker’s store I was presented with 15 or so different tablets of which nearly 1/3 were all single band. Then on to phones… While we know the biggest, baddest, newest phones out there all support 802.11ac, budget conscious consumers (the kind that will be bringing their phones to work) are being sold on insanely cheap deals for phones that are good enough to do what they need… and single-band.

 

You may have a corner case network where all your clients are brand spankin’ new, top-of-the-line devices and you can definitely make the shift to a 100% 5GHz WLAN but don’t expect the AP manufacturers to build just for you. Sadly, until the de facto is for all devices across the board to carry 5GHz radios and all drivers across the board to favor 5GHz I don’t see this happening industry-wide any time soon. If we as the experts want this to happen faster, urge your purchasing departments to stop cheaping out in the $15 upgrade, educate your users so they can make informed purchasing decisions, and keep fighting the good fight.

3 comments
4 views

Comments

Nov 18, 2014 02:49 PM

I wrote a whitepaper on this topic and posted it to my website at DivDyn.net just this week. While it will certainly take a while for the 2.4GHz clients to fade away, that doesn't mean that we have to design around them. Designing (or re-designing) WiFi networks can have a transitional mechanism (eg having a single SSID on 2.4GHz just for those legacy (or crappy) 2.4GHz-only devices. As a network manager you have to make the conscious choice to move away from 2.4GHz. I design around 5GHz and getting rid of 2.4GHz at the same time.

Devinator

Oct 31, 2014 05:49 PM

I saw that too. It's going to be such a long time because when the chip manufacturers stop making them, the surplus will hit the market. Once the surplus runs out, there'll be a used market for devices. Once that market fades, there will still be those few users hanging on to their old devices just because...

Unless we can petition the FCC to trade us the 2.4GHz band for more 5GHz channels I think my kids will be in the same situation.

Oct 31, 2014 09:45 AM

To get twitter back in the conversation, I saw a post that said. ( not quoted exact)

 

"We are still seeing B devices 15 years later, 2.4 will never die." <------- I smirked and muttered "true" when read it. 

 

Using that premise it will be 2030 before the 5GHz band even starts to feel any weight.