Wi-Fi was never a substitute for wireless carriers

There is a pervasive myth that Wi-Fi (802.11 wireless Ethernet) is some sort of substitute for mobile data services when the reality is that they are complementary technologies.  Wi-Fi is designed as a short-range technology while long range mobile standards like 3G CDMA or GSM are designed to blanket towns and cities with data coverage.  The limited range of Wi-Fi is a design choice that favors very high reuse of spectrum at the expense of range.  The lack of range is a necessary trade off yet it’s often and mistakenly cited as a liability.  The limited range is precisely what allows more people to reuse the same spectrum and do so license free and without coordination.

Note: Even at the mobile carrier level for 3G and 4G networks, reduction of transmission power directly correlates to higher capacity because it allows more cell towers to reuse the same frequency within a given area.  For example; 4 cell towers can operate in the same space as of one cell tower with twice the transmit power and that increases network capacity by 4.  The lower the range, the higher the capacity but higher capacity corresponds to higher deployment costs because more cell towers cost more money.

This misunderstanding of technology is why all the attempts to make Wi-Fi behave more like a mobile network (e.g., Muni Wi-Fi projects) have failed miserably.  Those projects were doomed to failure because they chose the wrong tool for the job.  But instead of acknowledging the fact that Wi-Fi or any system based on non-exclusive spectrum won’t work well for mobile networks, proponents of muni Wi-Fi double down on bad logic and argue that it was because Wi-Fi is saddled with junk spectrum and/or because it is too severely limited in power level.

These Wi-Fi spectrum myths are generally propagated by the spectrum commons advocates (AKA Open Spectrum) who believe that the radio communications should be mostly or completely unlicensed.  They argue that unlicensed wireless networking in the “white spaces” spectrum (the gaps between the TV bands between 54 MHz to 806 MHz) is a “Wi-Fi on steroids” due to superior propagation when it will be more like Wi-Fi on roid rage.  They argue that more (or all) spectrum should be left unlicensed instead of being auctioned off to carriers for exclusive use because that will leave more room for the type of innovation we see in Wi-Fi, but Wi-Fi innovation was never applicable to mobile Internet to begin with.  Wi-Fi innovation allowed us to be untethered from the nearest wall but not untethered from some kind of mobile carrier.

2.4 GHz spectrum junk status is due to the unlicensed nature and not physics

This story on the Washington Post from Ezra Klein appears to have bought into many of these spectrum myths when it made the following assertions:

You’re probably reading this on junk. And I’m not talking about newsprint – industry woes aside, that’s high-quality stuff. But if you’re on a computer or an iPad, and you’re not plugged into an Internet jack in the wall? Junk, then. But it’s not your MacBook or your tablet that’s so crummy. It’s the spectrum it’s using.

[SNIP]

But not all spectrum is created equal. “Beachfront spectrum” is high-quality stuff. Lots of information can travel long distances on it without losing much data. But not all spectrum is so valuable.”

These are misinformed statements because the problems with Wi-Fi (especially at conventions and trade shows where journalists often have Wi-Fi problems) is the unlicensed nature of Wi-Fi and not the lack of “beach front spectrum”.  In fact, we can transmit the same amount of data per second using the same channel capacity on 2.4 GHz as we could on 400 MHz TV spectrum.  We can even transmit roughly the same distances with 2.4 GHz as 400 MHz if terrain wasn’t a factor.  2.4 GHz may not have the propagation characteristics to bend around terrain as well as 400 MHz, but we don’t want better propagation for unlicensed radio like Wi-Fi or White spaces because higher propagation of unlicensed transmissions leads to a higher probability of interference.

It’s already bad enough that I get some 2.4 GHz interference from 4 neighbors in my immediate surroundings, the last thing I need is higher power 2.4 GHz interference from dozens of neighbors or from the entire neighborhood within a mile radius.  If someone wanted to use up every available White Spaces and Wi-Fi channel for wireless surveillance cameras even if it’s an inefficient use of spectrum, they are allowed to do so under an unlicensed regime.  The only way to minimize interference in unlicensed environments is to limit propagation which is why the FCC (and every other industrialized nation) intentionally limits Wi-Fi and White Spaces to 4 watt Effective Radio Power.

Not only are the power levels of Wi-Fi already high enough, but properly designed high capacity Wi-Fi networks in businesses and organizations often use much lower power than the permissible limit.  It’s common to see enterprise deployments limit their radios to 5 milliwatts when they’re allowed to go a hundred times higher because they want to be able to reuse the spectrum in every room to maximize network capacity.

Who needs wireless carrier corporations?

The ultimate goal of the spectrum commons people simply want to kill off the wireless carriers and leave everything to the device makers and the end users to innovate.  Ezra Klein’s article highlights this line of thinking by bemoaning the 2.4 GHz “junk” spectrum given to Wi-Fi and why the unlicensed Wi-Fi model of innovation should be used to replace the carrier corporations.  Klein wrote:

“That same section of junk spectrum became the home for WiFi – a crucial, multibillion-dollar industry. A platform for massive technological innovation. A huge increase in quality of life.

[SNIP]

Some advocates want that (TV) spectrum – or at least a substantial portion of it – left unlicensed. Rather than using telecom corporations such as Verizon to buy off the current owners of the spectrum, they’d like to see the federal government take some of that spectrum back and preserve it as a public resource for the sort of innovation we can’t yet imagine and that the big corporations aren’t likely to pioneer.”

This line of thought highlights the fundamental problem with the spectrum commons movement.  What kind of network would we have if it we had no wireless carriers?  Can we honestly build an ad-hoc wireless mobility network without wireless carriers with just intelligent devices using unlicensed spectrum?   “Unimagined” is certainly a good description since no one has come up with a workable alternative to a carrier model for mobile networks.

The spectrum commons movement believes unlicensed mobile networks are possible with technologies like ad-hoc mesh networking, but these mesh networks have proven to be unworkable in production environments because they’re fundamentally flawed.  Even Wi-Fi relies on miniature base stations called “Access Points” much less mobile networks yet the spectrum commons advocates have wasted years on client-only Wi-Fi mesh networks.  Without those high-powered and optimally placed cell towers and their back-haul connections to a global network, wireless devices (which have to be limited in size and power levels due to battery life concerns) are relatively helpless.

Short of a government built network using tax payer dollars, the carriers are the only ones who will risk tens of billions of dollars to build the nationwide cell towers and the back-haul networks.  Without some assurances on exclusive interference-free spectrum, no carrier will build any significant infrastructure on unlicensed spectrum.  While the spectrum may cost a few billion dollars, spectrum is a onetime expenditure that is dwarfed by the cost of building a new wireless network and dwarfed by the cost of building, upgrading, and maintaining that network.  AT&T for example spent $17 billion upgrading their wireless network in 2009 and $19 billion upgrading in 2010.  If carriers were so unnecessary, we’d all be using free walkie talkies instead of cell phones.

[Cross-posted at Digital Society]

About George Ou

George Ou was a network engineer who built and designed wired network, wireless network, Internet, storage, security, and server infrastructure for various fortune 100 companies. He is also a Certified Information Systems Security Professional (CISSP #109250). He was Technical Director and Editor at Large at ZDNet.com and wrote one of their most popular blogs “Real World IT.” In 2008, he became a Senior Analyst at ITIF.org, and he currently writes for High Tech Forum