The 2.4 GHz Spectrum Congestion Problem and AP Form-Factors

2.4 GHz is a junk band...
2.4 GHz is interference ridden...
2.4 GHz is dead...

You've heard all of these dire warnings about microwaves, bluetooth accessories, cordless phones, baby monitors, lions and tigers and bears (oh my!), which will wreak havoc on your WLAN.

Well, I'm here to tell you that 2.4 GHz issues are self-inflicted. And by self-inflicted I mean that Wi-Fi is the root cause of your Wi-Fi not working well. Specifically your own Wi-Fi. Poorly designed networks with too many APs blasting out too many signals all stomping on one another.

I'm a stickler for proper design, and is a mantra that I've probably beaten to death by this point. One aspect of proper design is deploying the proper number of radios to meet capacity needs. It's something that I talk about in my presentations on this subject and tweet about as well. Too many radios actually degrades performance due to the negative effects of co-channel interference and airtime utilization on a shared channel. More is decidedly NOT better!

Most WLAN designs today require a significant amount of
2.4 GHz radios to be disabled.

Most WLAN designs today require a significant amount of 2.4 GHz radios to be disabled (25-50% are common). Now before anyone gets all upset, this obviously requires analysis and design for each WLAN and there are variations. However, people don't like to disable radios due to a psychological reaction to avoid "loss" which is wired into us as humans. We have evolved to minimize "perceived loss" (emphasis on perceived) even if it leads to irrational decisions. This is called the 'Sunk Cost Fallacy' and you can learn more about it on the 'You are not so smart' podcast episode 35:

I blame the 2.4 GHz congestion problem
on the legacy fixed-band dual-radio AP form-factor.

I blame the 2.4 GHz congestion problem on the legacy fixed-band dual-radio AP form-factor. These are APs with two radios, where one is locked to 2.4 GHz and one to 5 GHz. A little stroll down Wi-Fi history lane... you see we only had single-radio 2.4 GHz access points early on. Then this fancy (but expensive) OFDM technology came out with 802.11a. Some vendors chose to start providing modular APs to add an optional second 5 GHz radio while still allowing the majority of customers to save money up-front by skipping 5 GHz and upgrading later if necessary (for instance, if they were one of the few who actually had 5 GHz 802.11a clients). Then silicon prices fell, 5 GHz clients became more prevalent, and 5 GHz radios in APs became cost-effective and desired enough that the industry settled on the dual-radio form-factor. This was a fairly major development in the industry.

However, having a fixed-band 2.4 GHz radio inside every AP has outlived usefulness and is now a detriment to modern WLAN performance. Client density increases and WLAN capacity demands have accelerated since the release of 802.11n in 2009. This has cascaded into the classic "throw more bandwidth at the problem" reaction by inexperienced or unknowledgeable IT personnel. In this case it manifests as "throw more APs as the problem," with the false thinking that simply more APs result in more capacity.

You see where I'm heading here. More APs with more fixed-band 2.4 GHz radios coupled with a human instinctual reaction to avoid perceived loss leads to a default state of congestion in 2.4 GHz. You have to be comfortable with and accept disabling 2.4 GHz radios to be successful today.

The future is software-defined...
Software Defined Radios (SDR)

Vendors can help reduce the congestion problem by developing APs that are more flexible and avoid the psychological sunk cost problem all-together. The future is software defined... Software Defined Radios (SDR).

APs with more flexible options and band unlocked radios are on the horizon. This is sometimes also called software-defined radios (SDR) (although there are distinctions that are beyond the scope of this post). Vendors can help alleviate 2.4 GHz spectrum congestion by eliminating the perceived loss that comes with unused radios. They have made minor sales and marketing attempts at this already, telling customers that unused 2.4 GHz radios don't need to be disabled but can provide value as monitor mode radios providing full-time intrusion detection, spectrum analysis, remote packet capture ability, and enhanced location tracking. However, these arguments have been less than convincing, with customers still feeling like they're not getting the value they're paying for (that perceived loss kicking into effect). And attempts at selling single-radio 5 GHz APs aren't really attractive from a financial perspective, because the cost of the AP is only minimally reduced compared to a dual-radio design. 

More flexible AP form-factors will become more prevalent in the market to address the need for greater WLAN capacity, take advantage of more 5 GHz spectrum availability, and introduce ultra-high speed in-room performance with the 60 GHz band and 802.11ad. Some vendors have been playing with the technical feasibility of band-unlocked radios and multiple radios in adjacent bands within very close proximity (inside the same AP) already. This is tricky stuff to avoid interference inside the AP, requiring specialized skills and custom circuit board design. Not all vendors have the chops or expertise to pull this off, so we'll likely see the market share leaders drive this as a custom-designed and differentiated product initially, with smaller market vendors trailing once reference designs become available from ODMs (Original Design Manufacturers; many smaller Wi-Fi vendors don't actually design their own APs but take stock reference designs and tweak them if anything). I believe vendors will work all these complexities out, if they haven't already.

A dual-radio AP, with one radio band-unlocked (2.4 GHz or 5 GHz)
and one radio fixed in 5 GHz, is likely the “sweet spot”

We will likely see APs in various form-factors. A dual-radio AP, with one radio band-unlocked (2.4 GHz or 5 GHz) and one radio fixed in 5 GHz is likely the easiest and most cost-effective option for vendors to deliver. This also has potential to be the "sweet-spot" for the majority of APs being sold in the near future. Tri-radio APs providing an additional 60 GHz band radio should also appear but will be higher cost and likely a solution for spot-use where ultra-high speed or density is required beyond what 5 GHz can provide.