I got a lot of grief for suggesting in a Wall Street Journal op-ed "It's Time to Shut Down the FCC" that there is no scarcity of electromagnetic spectrum. Here are some of the comments the WSJ received:
- Mr. Kessler should stick to hedge funds. He clearly doesn't understand the physics involved in wide-area wireless technology. His statement that there is no scarcity of wireless spectrum illustrates his lack of expertise in this area.
- Unfortunately much of Mr. Kessler’s comments are based on false premises. Bandwidth is actually very finite - go back and check freshman physics - and the FCC has really struggled to have more allocated from the old VHF TV bands and has considered options for the UHF TV bands.
- Of course there is spectrum scarcity. Two signals can't operate the same space and time without disrupting each other. But the laws of physics interferes with desires of Wall Street wheeler and dealers to make a killing by selling the same thing twice, or even three times.
- Bandwidth is an extremely scarce resource -- I don't know anyone in the business who would not agree with that basic statement.
It really stings (OK, actually its humorous) to be told I don't understand physics, especially after taking so many damn physics courses to complete my electrical engineering degrees. But that was a long time ago, maybe physics has changed. Here are my egregious statements.
- "Founded in 1934 partly to regulate radio spectrum (which in reality hasn't ever been scarce), the FCC..."
- "During the last two decades, the FCC has manufactured the idea that the electromagnetic spectrum used by wireless devices is scarce."
- "The FCC can apply more Band-Aids to its broken scarcity-based regulatory model. Or we can close the agency and let consumers allocate capital efficiently."
- "There is plenty of bandwidth, both inside of fiber optic cables and out in the air, and more everyday as faster chips can handle higher and higher speeds."
Look at this chart of the airwaves. Fun colors and everything. Click and you'll get a high res PDF of US Spectrum Allocation.
For the most part, your smartphone uses spectrum in these bands 824–849 MHz, 869–894 MHz, 1850–1910 MHz, and 1930–1990 MHz. It's a tiny sliver in the 5th row down labeled 300 MHz to 3 GHz.
Note that a huge chunk of spectrum in one row is 1/10th the size in the next row. So while the 160 MHz of spectrum seems like a lot in the 18/1900 MHz (1.8 and 1.9GHz) band, its really tiny in the 18 GHz and 19 GHz band.)
One of the ways that operators do get more out of existing spectrum frequencies is they break geography into cells (hence the name). You can squeeze more out of alleged scarce spectrum by going to smaller and smaller cell sizes. You can't shrink them forever, but you can for a very long time.
But there is plenty of spectrum above 3 GHz and even more above 30 GHz. Now, I'm not making my case for a lack of spectrum scarcity based on misallocation of spectrum by the FCC (though that's certainly true), but instead by looking at the next two rows. At around 5 GHz there is spectrum used for Wi-Fi and cordless phones and that kind of stuff. It was completely unusuable a decade or so ago, until faster chips were mass produced to enable commercially available products. Was that new spectrum? No, of course not, it was always there. Because of advancing technology, spectrum scarcity, if it exists at all, is always fleeting.
As you go up in frequency, the wavelength gets shorter meaning the waves don't travel as far before degrading (see, even though I sat in the back of the lecture hall, I still paid attention!) This means making cell sizes smaller. While a problem for the current AT&T and Verizon Wireless business models of utilizing big ugly tall cell towers , it doesn't mean its not possible or even economic. AT&T will even sell you a femtocell for your home (they should give you one since it covers up the shortcomings of their big ugly towers.)
But technology has always advanced to solve any fleeting scarcity issues. Semiconductor technology exists today for affordable handheld devices to operate in the 5-10 GHz range, and maybe even up to 30 GHz. This was my point with the "There is plenty of bandwidth...and more everyday as faster chips can handle higher and higher speeds" line. There is AT LEAST ten times as much room for high speed data as there is at 1900 MHz.
But even 30 GHz is nothing - we can't stop there. Look at this paper from NTT in Japan. It describes a demodulator capable of operating in the 120 GHz band. The chip is capable of 10 gigabits per second speeds. Your 3G cellphone can maybe do a megabit per second (after an annoying one second delay.) The NTT chip is surely uneconomic today, its just a sample chip in a lab somewhere in Tokyo. But that high speed chip in your phone today was in a lab ten years ago, maybe less. And that microprocessor in your PC wasn't even buildable ten and maybe even five years ago. Yet you have it today and at a price point so cheap you'll throw it away in another 18 months. That's how fast technology is advancing.
When I claim there is no spectrum scarcity, all I am doing is looking at the map of INFINITE frequencies and plotting the forward technology curve that will make bigger and bigger swaths of that colorful spectrum allocation chart usable in mass produced commercially viable products. The FCC, almost by definition, takes a static view of the airwaves. Markets take a dynamic forward looking view. Which would you rather trust?
