Interesting People mailing list archives
Re: It's Silicon Valley vs. Telcos in Battle for Wireless Spectrum
From: David Farber <dave () farber net>
Date: Wed, 16 May 2007 20:38:50 -0400
Begin forwarded message: From: Bob Frankston <Bob19-0501 () bobf frankston com> Date: May 16, 2007 2:09:02 PM EDTTo: "'David P. Reed'" <dpreed () reed com>, "'Bob Frankston'" <bob37-2 () bobf frankston com> Cc: dave () farber net, ip () v2 listbox com, "'Dewayne Hendricks'" <dewayne () dandin com> Subject: RE: [IP] It's Silicon Valley vs. Telcos in Battle for Wireless Spectrum
If you want to carry a television signal over a long distance then use that darn Internet. The problem is that we have layers upon layers of simplifying assumptions and we pile them on instead of rethinking. This is a key point
in Robert Laughlin's "A Different Universe" (a DPR recommendation). We tend to solve problems as if they exist in isolation. We do need to decompose problems in order to deal with them but there isn't a unique decomposition and we have to be open to rethinking the decomposition. In the early 1900's the first decomposition was to assume that"communications" is a unit. We then decomposed that into voice and video. We also had wired vs wireless. The accidental properties of wires led us to use
them for one-to-one communications because they were point to point connections. Wireless signals couldn't be contained hence they werebroadcast to all. It doesn't help that "radio" and "telephony" are now used to describe business rather than technologies thus making it hard to talk about the basic assumptions separate from the market assumptions. This is
why TV over wires is not treated as a point to point medium -- it wasoriginally implemented as shared antenna system and our policies continue to
make that implicit assumption. Today this decomposition is extremely dysfunction. We now understand the concept of bits as a common representation for information and that is avery effective point of decoupling as we've seen with the Internet. We can
then use packets to organize the bit transports.This decoupling of the communications from the bit transport (as in TCP/UDP vs IP) is very hard for people to accept because we base our understanding
on the naïve assumption that we are simply relaying sound waves through electronic tubes. There is assumption that you must preserve therelationships between the bits as in isochronous and QoS models. This was a defining assumption for analog telephony and the solution, eventually, was digital telephony and then packet telephony. We treated the answer as way to
make analog telephony work better without using the new understanding to question whether we were asking the right question in light of our understanding. Digital technology freed us from the constraint of isochronicity and thus allowed us to get many orders of magnitude inimprovement (by various measures). But it was an answer the put a lie to the decomposition that defined the industry that asked the question in the first
place and thus they had a stake in failing to understand the answer they got.Once you decouple the bits from the interpretation you can view wired and wireless bits indifferently and can dispense with the complexity and expense of slicing and dicing the transports and the use of special kinds of wires
and gear for each particular message.Yet we continue to argue as if it is still 1934 and every bit is special -
sort of like the $100 Japanese Honeydew Melons but far pricier.These concepts extend far beyond "telecom" into our basic understanding of
how systems work. The isochronous assumption has a parallel in the assumption that we have to solve problems as stated and that if any component fails the system fails. Thus the Y2K scare and the presumptionthat we must govern systems lest people do things the "wrong" way -- alas wrong often means finding that we've solved far more interesting problems --
we may have wanted a guidance system for airplanes but we got digital computers instead. We presume music comes from record companies and notmusicians and thus we preserve a particular industry structure rather than
allowing for musicians to be heard.Telecom is useful case study because it's a simple problem and the price of
continuing to live in 1900 is higher than we should have to bear. Too bad this fight over slivers of spectrum is treated as if it were anymore real than the rest of professional wrestling. The entertainment value
doesn't make up for the collateral damage. -----Original Message----- From: David P. Reed [mailto:dpreed () reed com] Sent: Wednesday, May 16, 2007 12:40 To: Bob Frankston Cc: dave () farber net; ip () v2 listbox com; Dewayne Hendricks Subject: Re: [IP] It's Silicon Valley vs. Telcos in Battle for Wireless Spectrum To amplify Bob: The standard term for unhooking functions from transport is "transport independence". There is no fundamental reason whatever that television pictures work best over the VHF portion of the spectrum, for example. (Please don't start yet another debate about how VHF goes longer distances - I won't engage unless you can tell me how much energy is in a VHF photon or how that relates to its diffusion through multiscale structures of building and living materials and its diffraction through the fresnel zone of the earth). There is no spectrum to liberate. There is no real object called the spectrum. There *are* allocations of rights to radiate certain waveforms - in other words, collections of words on paper that control the legality of actions that involve applying power to collections of electronic gear that create electromagnetic waves - and those allocations are phrased in terms of wavelengths or frequencies and maximum power at certain locations within the US airspace. Those allocations of permission (so-called license rights) are indeed scarce. They are scarce by fiat. There is a pretense that they must be scarce because of laws of physics. However, the laws of physics say no such thing. The information carrying capacity of the electromagnetic field is vastly underutilized as a result of this badly designed system. It was a good system in 1934 because engineering barely knew how to use the field and its waves. It is not a good system today, knowing what we know. And there is a lot more that we could know, if we could invest in improving the system as engineers, rather than being stopped by ignorance on the part of lawyers, politicians and "policy experts". Bob Frankston wrote:
This is portrayed as a battle between those who want to "liberate" slivers of spectrum and those who want to keep control or, to put it another way, about who will divvy up the spoils. We're so mesmerized by the theatre of this epic battle that we don't ask whether spectrum allocation makes senseeither as a basic concept or the way we approach it. This spoils model provides fodder for Cleland to call openness closedness. Even if think spectrum allocation is necessary for technical reasons
(though
as David Reed and others point out, it's not) tying bits to the pathminimizes the value by designating particular service providers as ownersrather than providing common paths for bits. It's also annoying that that the story itself shows a lack of
understanding
of the cellular industry. One of the major "advantages" of CDMA is to fixthe bug in GSM that gave the user an even break -- you must indeed use instruments approved by carriers in order for the carriers to deign to
allow
you to use "their" network. As I've pointed out, just because my phone litup four bars in Australia, Telstra wasn't going to annoy Verizon by
letting
me use my phone.The article doesn't seem to show an awareness that a GSM phone fATT (fauxATT) would work just as well on T-Mobile. It's likely that fATT will disabled this to prevent the phones from being used on other networks. There's also the question of what protocols the phone uses for its
services
and whether they are network-independent. I don't know the answers thoughhope to learn them soon. Too bad the article is fixated on the theater of the wrestling match
rather
than providing or solid information about the particulars and about the larger issues. No matter who gets these slivers of spectrum we'll all belosers. -----Original Message----- From: David Farber [mailto:dave () farber net] Sent: Wednesday, May 16, 2007 09:26 To: ip () v2 listbox com Subject: [IP] It's Silicon Valley vs. Telcos in Battle for Wireless
Spectrum
Begin forwarded message: From: dewayne () warpspeed com (Dewayne Hendricks) Date: May 16, 2007 3:58:14 AM EDT To: Dewayne-Net Technology List <xyzzy () warpspeed com> Subject: [Dewayne-Net] It's Silicon Valley vs. Telcos in Battle for Wireless Spectrum [Note: This item comes from reader Randy Burge. DLH] It's Silicon Valley vs. Telcos in Battle for Wireless Spectrum Frank Rose Email 05.16.07 | 2:00 AM <http://www.wired.com/techbiz/media/news/2007/05/uhf_spectrum> Apple's iPhone may be the most eagerly awaited gadget of the year, but when it finally goes on sale some time next month, only 30 percent of US mobile phone customers -- those who subscribe to AT&T's wireless service -- will be able to use it. Verizon subscribers might have had a shot, but executives at that carrier nixed the idea of letting an Apple device onto their network years ago. It's as if Mac owners had to connect to the internet through AT&T because their machines wouldn't work on Verizon, Comcast or Time Warner Cable. The wire-line internet doesn't work that way, and wireless doesn't have to either. By the end of this year, the FCC is expected to start auctioning a frequency band that could be used for a wireless network that any device -- be it a cell phone, laptop, desktop, TV or toaster -- would be able to connect to. A proposal to build such a network has been presented by Frontline Wireless, a startup backed by three of Silicon Valley's biggest players: Venture capitalist John Doerr, Google angel investor Ram Sriram and one-time Netscape CEO James Barksdale. But Frontline will be bidding against behemoths like Cingular and Verizon, and whether it has a chance will be determined within the next few weeks, when the FCC sets the rules for the auction. The spectrum that's coming up for grabs is prime stuff: A large, low- frequency band that's currently being used by UHF television stations, which have been ordered to vacate it when broadcasting goes digital in February 2009. UHF may not be as good as VHF, which operates on even lower-frequency spectrum. But it has the ability to carry information through forests, buildings, even mountains, regardless of the weather, and that makes it ideal for broadband wireless, or for mobile-phone service. Ever wonder why Cingular and Verizon, the biggest and most successful U.S. carriers, can offer more reliable service than Sprint or T-Mobile? Because the big boys already own a large band of spectrum near the UHF band, while the little guys are stuck with spectrum that operates at double the frequency and is far less powerful as a result. <snip> ------------------------------------------- Archives: http://v2.listbox.com/member/archive/247/=now RSS Feed: http://v2.listbox.com/member/archive/rss/247/ Powered by Listbox: http://www.listbox.com
------------------------------------------- Archives: http://v2.listbox.com/member/archive/247/=now RSS Feed: http://v2.listbox.com/member/archive/rss/247/ Powered by Listbox: http://www.listbox.com
Current thread:
- It's Silicon Valley vs. Telcos in Battle for Wireless Spectrum David Farber (May 16)
- <Possible follow-ups>
- Re: It's Silicon Valley vs. Telcos in Battle for Wireless Spectrum David Farber (May 16)
- Re: It's Silicon Valley vs. Telcos in Battle for Wireless Spectrum David Farber (May 25)