It's not just open standards, it's open spectrum

While not often disagreeing with Groklaw, I did have a concern with this piece on open standards and disaster relief. While agreeing with the general gist — that proprietary standards can hinder relief efforts and that open platforms can maximize the number of people helped — the problem of interconnecting communications services is not generally an open standards problem. It is an open spectrum problem — and one we could solve today.

Ever since 9/11, the communications world has known about the problems of interconnectivity among public safety services and the general public. Long ago, when radio was becoming a useful tool for public safety and not just for broadcasting, the FCC divided up the spectrum into little slices and gave them out _by service_. Thus, “firefighters” got a slice of spectrum, “police” got another, “forest rangers” a third, etc.

This seems nuts today, but at the time it made sense. Most of the time, firefighters and police officers don’t need to communicate by radio. In the old days of primitive radios and powerful transmitters, there was fear that police radios and firefighter radios on the same band would interfere with each other. So the FCC put them on seperate bands and relied on the stationary command centers (the fire house and the police station) to communicate with each other and then retransmit to the units of their forces that needed to coordinate. While this did not allow instantaneous coordination between units on the scene, this was felt to be a reasonable trade off against the risk of regular interference, police responding to fire calls, etc.

Unfortunately, we are now living in a world where each service has its own special type of highly expensive (because economies of scale never kick in) radio on its own frequency. It’s not just open standards, the radios simply cannot communicate with each other because they do not send or receive on the same frequencies.

The second problem is overwhelming existing infrastructure by civilians and triaging civilian use with emergency use. Even in 9/11, when cell sites (other than those on the Towers) were not destroyed, the sheer volume of civilians trying to call each other overwhelmed the system (although text-messaging, which uses much less bandwidth, worked better). Public safety folks want a way to take over that bandwidth if needed. At the same time, however, you don’t want to cut off the civilian population completely from telecom services. Civilians need to coordinate with each other to ease evacuation, locate missing relatives, or even report new developments to authorities.

Finally, there is the problem of replacing critical infrastructure when it gets destroyed. Radio isn’t magic, and communications systems designed around a particular architecture, like cellular, will not work with a different sort of architecture.

Unfortunately, the solutions offered do not seek to leverage advances in communication technologies. They are astoundingly conventional “more of the same” answers, aided and abetted by the manufacturers who want to continue to make and sell high-priced equipment with nice profit margins. The public safety community wants a slice of the analog television spectrum (when returned) for a national public safety band to which existing services would migrate. Post 9/11, the FCC convened a working group of teh wireless phone comapnies to develop ways public safety could take over capacity on the cellphone networks (to the exclusion of civilian traffic).

What’s frustrating is that we already _have_ a good technical solution. For years, open spectrum advocates have touted the virtues of “smart radios.” Using existing technology, a radio can scan the spectrum for usable bands, find the receivers that should receive the messages, and use the unused bands to send the message. The receivers ignore incoming signals unless they are tagged to attract attention, in which case they tune in.

To see how this plays out, lets go back to the firemen and police and national guard and power companies and telephone folks all struggling in the aftermath of Katrina, while civilians desperately try to find each other and needed services. Lets give them each smart radios that have different sets of programming.

The emergency responders have radio set to find each other in the same way your cell phone waits to receive a call. Their signals are also tagged with a priority signal that allows them to “bump” lower priority signals (we can give the military an “uber user” privilege). Shared encryption ensures that no one can easily evesdrop. The rules also allow them to sense if desirable (because of penetration characteristics) television or FM radio bands are unused and allow them to use those bands. Ad hoc networks in mobile units use higher power antennas to set up citywide networks and provide backhaul.

At the same time, “smart” cell phones receive an emergency code permission to use any available carrier with whatever rules are determined by relevant safety authorities to resolve increased demand and provide access to public safety. In fact, cell phones could be designed to receive calls or text messages bearing a public safety stamp, so that needed infromation (e.g., the Main Street entrance to the highway is flooded) can go to individuals who need it and might not have access to radio or television.

As an added bonus, this would also drive the price of equipment down enormously. If everyone can use smart radios, economies of scale will kick in and the price will plunge dramartically. We saw an example of this last year. Public safety now has a band in 4.9 GHz. The public safety orgs petitioned the FCC for permission to have a rules change that would allow them to use wifi chips on the frequency (chips designed for 5 GHz can be tweaked via software to work on 4.9 GHz). Motorolla objected, arguing that this would create interference in the band. The public safety folks responded that using wifi chips would drive the cost per unit down from about $2000 per unit to about $20 per unit, purely as a function of economies of scale. (The FCC granted the petitions, subject to lower power use for the “wifi” units to address the interference issues.)

So what’s the problem? The FCC’s rules prohibit all but the most primitive “smart radios.” The FCC prohibits the sort of frequency hopping described above. In fact, to get a device approved for operation, you need to show the FCC’s engineers that the device will not broadcast out of band, and that you have taken steps to prevent others from tampering with the device to make it operate out of its assigned band.

Why? Because incumbents object to smart radios moving accross “their” bands even if no one is using the band and primary licensed services could displace anyone else. For some, this is an interference question: what if the smart radio technology doesn’t work as advertised and thousands of supposedly “smart” cell phones interfere with my authorized service? All the engineering data in the world can’t convince some of these folks that what they learned in school 20 years ago might change as a result of technology. They don’t trust it, and gain nothing by seeing other people use “their” spectrum, so why take the risk?

But there is also an economic aspect. Once radios can move in any spectrum and do anything, you no longer need one of the few licenses in a metro area to provide cell phone service to be a cell phone operator, or need a TV license to be a TV station, etc. Any spectrum will work, because the smart radio will seek out the content it wants. In fact, you might get rid of exclusive licensing at all, since if smart transmitters avoid interfering with each other, than why have exclusive licensing?

Needless to say, this is one of my big policy fights and why the issues of open spectrum matter so much.

Of course, we would also need open standards of some kind to make the system work, and we want open standards to avoid the idiocy of a FEMA form that you can only fill out with one type of browser. But open standards alone will not solve the communications problems. It is an open spectrum problem, and one we could solve today rather than wait for the conversion to digital television.

Stay tuned . . .


  1. Wow,

    That is a fascinating article, and very well explained.

    You and I REALLY need to get back to work on that draft article we started at Worldcon. And when I say “You and I,” I mean of course, “I”, since I’m the person gumming up the works.

    Stay tuned. . . soon, I hope.

  2. This is a critical issue that demands attention and action. I attended a telecommunication forum that Rep. Jay Inslee sponsored in Seattle earlier this month. Sadly, Inslee believes that 70 million “stranded” analog TVs are a major obstacle to spectrum reform and the Digital TV transition. This is a bogus argument pushed by broadcasters . . I have a longer rant on this at http://blog.seattletimes.nw

    but thank you Harlod, for backing me up.

    Bart Preecs

  3. Harold–

    I like this very much. What do you think of requiring that public safety bands be compatible with Wi-Fi standards, as was done at 4.9 GHz over Motorola’s objections?


  4. I’d need to know a little bit more about the current set of chips used for public safety and the frequencies on which they operate. In 4.9, there was a question about emission mask and spurious emissions into neighboring bands because the chips designed for 5 gig are in a different standard. That’s why you can only use the “wifi chips” at lower power. I think a fair amount of public safety use is in crowded spectrum below 1 Gig, and I’m not sure the current emission mask for chips using 900 MHz PArt 15 spectrum is useful. Gotta put the question to a real engineer.

  5. Which public interest group is carrying the torch on this right now (or should be). PK? EFF?

  6. That would be Media Access Project and, more specifically, me. It’s my day job.

    PK says they care, but I have yet to get them real excited on actually working on this.

    EFF used to care, but hasn’t really been a player since Cory Doctorow moved to England to do his other projects.

    The major groups that do federal spectrum policy are MAP, CUWIN, New America Foundation (NAF), Free Press, Consumer Federation of America, and Consumers Union. EDUCAUSE has been more involved recently. Others, such as CTCNet, are usually good for comments when prompted (such as on the recent 3650-3700 MHz proceeding).

    A great frustration has been getting the broader open source community to care about the policy. The GNU Radio project, for example, encourages people to write code, but does nothing to help organize for changes in spectrum. Slashdot, the great news aggregator for the open source community, does not cover this stuff at all.

    If we can ever get the cats to all come together in one pride and sing off the same hymnal, it’ll be a roar to wake the dead and move mountains. But outside of Sascha, I can’t get anyone in the open source community to care.

    Even so, we haven’t done too badly. The problem is we have come to the end of the free ride. The reforms Powell started have wound their way to conclusion. Martin has little interest in this. It will take significant coordinated polictical pressure to make anything else happen from now on.

    I’m planning on spending a lot of time talking about this at Esme’s conference. I keep meaning to put a spectrum policy page up on your website, but I haven’t had time yet.

Comments are closed