There’s a lively discussion about the DTV transition going on right now. (Go ahead, ask an elderly neighbor of yours who receives television broadcasts over the air what she knows about the DTV transition. I bet not much.)
John Kneuer of NTIA recently testified that the special single-purpose converters are indeed being manufactured, and that IBM is doing a fine job organizing the entire coupon program. Of course, it looks like a mess from the outside (and once Kneuer finished testifying he left NTIA – a year before he was expected to), but maybe this will turn out to be another Y2K non-event given all the coupon-device planning that has taken place.
But GAO is saying (and Commr. Adelstein is saying, that someone really should be in charge.
Here’s the thing: if a master coordinating plan is developed, it wouldn’t possibly be implementable by … January 2008, when the coupon plan is supposed to kick in. That’s right, two weeks from now. So there would need to be a delay. What possible consequences from such a delay? Well, how about slowing down the auctioning of that 700 MHz spectrum. No transition, no hope of cleared spectrum, all the assumptions of the bidders go awry. That couldn’t happen, could it?
Meanwhile, last week in the white spaces:
1. Philips has submitted its own device for testing, with an improved user interface based on feedback from FCC engineers.
2. Motorola suggests that to protect incumbent broadcast and cable signals, portable devices have to be REALLY low power. Motorola has tested some really old, worn-out cables and shows they may be susceptible to interference.
Motorola wants opportunistic, signal-sensing sorts of devices to be available only for home use, and suggests that higher-power outside-home uses be driven by geo-location database registration information (in other words, the incumbent says in a database what its spectrum contour is, and that’s taken as gospel by the device no matter what different result sensing might lead to – and the device itself has to register with the database).
Motorola recommends that Class B consumer devices be limited to 10 mW to minimize the potential for in-home interference. We also recommend that devices outside the home environment be allowed to operate at power levels up to 4 W EIRP, along with requirements for geolocation, sensing and support for a beacon. (Designated as Class A devices in the prior white paper.) This multi-tiered classification provides for high power devices needed when deploying broadband to rural areas and responding to the growing broadband mobile or portable needs of professional users. Use of a geolocation database in combination with registration requirements will help ensure that these applications are for professional or outdoor uses, and provide additional mechanisms to ensure interference is not caused.
…The experimental Motorola WSD [for use outside the home] utilizes geo-location database techniques in order to determine available spectrum and the maximum allowed transmit power levels per channel (vs. location). The unit does not rely on sensing techniques to determine open TVWS spectrum (i.e., to determine if the unit is inside of a TV station’s protected service contour) – it relies on geo-location database techniques for such information.
The problem with this approach is that it’s hard to build a business based on devices that are so low power you have to be sitting on the lap of the signal to receive or transmit. And the real market here is mobile use – anytime, anywhere.
3. Microsoft and Philips are going to do some more testing, and talked about wireless mics. I understand that a huge percentage (80%?) of wireless mics operate illegally.
4. Google has its own devices that it wants to have the Commission test, and they’re based on repurposed equipment. (It was probably a little uncomfortable for Google to have to rely on the secretive Microsoft in the testing regime – better, probably, to go it alone.) I remember seeing the historic Google server at the Computer History Museum, with its sagging repurposed racks.Cheap equipment used redundantly can be enormously successful.Here’s the Google report:
The Google representatives presented a demonstration of preliminary test results, based on the initial phase of ongoing trials involving two forms of experimental technology iutilizing repurposed equipment.In both instances, these test results demonstrate that digital televisions (DTVs) and wireless microphones can be amply protected from harmful interference by unlicensed personal/portable devices, using reasonable power levels and sensing thresholds.
What they’re doing is spread-spectrum sensing, I think.
Problem: Narrowband pilot-based schemes are subject to “deep fades”, making power estimation unreliable, requiring large margins
Broadband DTV sync signals give reliable power level estimates, even when the pilot is deeply faded, removing the need for large margins.
Unlicensed devices can safely coexist with licensed devices, without fear of harmful interference
Broadband sensing technologies can greatly improve the accuracy and reliability of spectrum sensing
Burst transmissions inherently cause less harmful interference to existing licensed devices.
All of this fits together. I think. As I understand it, you can’t do proper highspeed internet access using the 22 MHz of 700 MHz being auctioned off as Block C. Even if you had access to/control over the entire 60MHz portion available for commercial auction, it wouldn’t work. I understand you can do about 2 bits on one Herz. So 60 million Herz would equal 120mbit/s. But this would be a shared medium, so you’d be stuck with at most, for ten households, 10mbit/s down and 2mbit/s up. Not so zippy.
That’s why the white spaces are so important. The 700 MHz auction sets the precedent for the entry of the internet ethos into the wireless carriers’ world, but it won’t solve the nation’s highspeed internet access problem. Opportunistic low-ish-power portable devices capable of operating on unlicensed white space spectrum could be used to avoid the last-mile bottleneck for internet access. They could allow for short hops in isolated places to shared fiber connections. FreePress takes the position that “[u]sing these white spaces, the wireless broadband industry could deliver Internet access to every American household at high speeds and low prices — for as little as $10 a month…” Cooperative neighborhood mesh networks could use the white spaces to share a single fiber connection to the internet with hundreds of people.
So that’s why tech giants and public interest groups have formed a coalition to support unlicensed, portable uses of the white spaces. If the 700 MHz auction gets screwed up by a failure to adequately plan for the DTV transition, the white spaces can still move forward – and could really help with highspeed internet access in underserved areas.
*Cross-posted from [Susan Crawford blog](http://scrawford.net/blog).*