WASHINGTON Broadcasters hammered the FCC's upbeat take on its white space tests, saying the hype doesn't match the actual data in the commission's own report.
At press time, Martin had a vote on his proposal—which will not become public until some time after a vote—on the agenda for the commission's meeting Nov. 4 (Election Day), the day Google founder Larry Page demanded the commission produce the rules.
Since the FCC released its report Oct. 13 and Martin outlined his plan—which no one outside the commission will actually see until some time after it's voted upon—NAB and others fought back. Oct. 17, NAB, the Association for Maximum Service Television and the four main networks filed an emergency petition with the FCC, calling for a public comment period on the proposal and a delay of the Nov. 4 vote.
GOOGLING THE PLAN
The far-reaching proposal would allow unlicensed devices in DTV white spaces—including on channels adjacent to occupied DTV channels—at power levels close to what was recommended by Google and other high-tech companies.
According to Martin's description, portable devices that use both spectrum-sensing technology and geolocation databases to prevent interference to DTV would be limited to 100 mW of power, or 40 mW when operating on channels adjacent to active DTV channels. If the devices use only spectrum sensing, they would be limited to 50 mW, or 40 mW on adjacent channels.
The Association for Maximum Service Television said that the 40 mW level on adjacent channels would "eviscerate" DTV in some areas, particularly in cities.
Fixed devices would operate at a maximum of 4 W, and not at all on adjacent channels unless they can show to the commission that they can avoid interfering with DTV.
But so far, the devices have had a tough time when DTV is on adjacent channels. "In some cases, the degradation [in detection sensitivity] was such that the detection threshold could not be measured," OET said in its report. "This could impact significantly the ability of the devices to reliably detect TV signals within stations' service areas."
That could block white space proponents' plans for any major hub-based applications, such as wireless broadband, in urban areas, because there are so few available channels if adjacent channels are off-limits. There will be even fewer channels available once Low-Power and Class A stations migrate to their final resting places among Channels 2-51.
Martin's power limits come close to those in a recent proposal by Google. Its regime would rely on geolocation, but not spectrum sensing. For non-adjacent channels, it proposed maximum power levels of 36 dBm (about 4 W) for fixed devices and 20 dBm (100 mW) for portable devices.
Broadcasters have attacked the reliability of spectrum-sensing technology, and the report notes the poor performance of some of the spectrum-sensing devices. But the OET says "proof of concept" was shown for spectrum sensing.
Broadcasters have also called for lower power levels. MSTV wants no fixed devices in adjacent channels and portable devices at just 5 mW on those channels and 10 mW on the other channels.
FAULTY INTELLIGENCE?
The OET report shows a mixed performance by the devices. On spectrum sensing, OET said in most cases, the devices correctly reported channels as occupied when the device was operated within the service contour of the stations broadcasting on those channels and signals were viewable.
But in some instances, devices from Adaptrum, the Singapore-based Institute for Infocomm Research (I2R) and Motorola (in sensing-only mode) incorrectly reported channels as unoccupied (available) even inside a station's service contour and when the signal was viewable.
Furthermore, there were many false positives. All of the devices reported some channels as occupied outside of the service contours of stations broadcasting on those channels whether the signal was viewable or not.
The device from Philips had the most false positives, said OET. It generally reported most channels occupied, whether operating inside or outside any service contours and whether the signal was viewable or not. It had similar problems with detecting wireless mics. "At both sites and all the test locations, the Philips device reported all the channels on which the microphones were designated to transmit as occupied whether the microphone was transmitting or not," OET reported.
Other devices were better at detecting wireless mic signals, but they suffered detection degradation in the presence of DTV signals in adjacent channels.
MSTV said the technology was not ready for primetime.
"These devices failed to differentiate between an occupied and unoccupied TV channel nearly 33 percent of the time.
Specifically, our analysis of the data reveals these devices have a 'sensing error rate' ranging from 26.9 percent to 37.3 percent," MSTV said in a statement. "Allowing millions of unlicensed devices, which rely exclusively on sensing to avoid interference, on to TV channels will decimate digital television reception across the country."
NAB said the discrepancies between the OET data and its own report were "curious" and contrasts "stark."
"It would appear that the FCC is misinterpreting the actual data collected by their own engineers," said NAB Executive Vice President Dennis Wharton in a statement. "Any reasonable analysis of the OET report would conclude that unlicensed devices that rely solely on spectrum sensing threaten the viability of clear television reception. Basing public policy on an imprecise Cliffs Notes version of a 149-page report raises troubling questions."
CONSERVATIVE ROLLOUT
Motorola got a thumbs-up for its use of a geolocation database in combination with spectrum sensing. "In those tests, the Motorola device correctly reported all occupied channels used by stations within whose contours the WSD was operated," OET said.
Motorola Senior Director for Regulatory and Spectrum Policy Steve Sharkey said the report, from what he'd seen, was "extremely positive," although the public won't see any more details of the plan until after it's voted on.
"It really affirms what we've been saying, that the devices can protect the incumbents," he said.
He predicted the FCC, concerned about interference, will be conservative with the devices allowed, but reach a higher comfort level as manufacturers develop and test more devices.
He said the rules will probably be sufficient to allow manufacturers to actually start building devices. The kind of devices and service may vary wildly from company to company, ranging from in-home networking to rural wide-area wireless broadband.
Meanwhile, standards development for the channels continues. IEEE has been working on such standards since the FCC began the current rulemaking back in 2004. Winston Caldwell of the Fox Technology Group, at the IEEE Broadcast Technology Society symposium in October in Alexandria, Va., gave an update on the proposed technical requirements of such a system.
He said the service would provide broadband on unused VHF and UHF channels for residents of rural and outer suburban areas and listed a number safeguards that are woven into the proposed standard (IEEE 802.22) on wireless regional area networks (WRANs) to protect television broadcast operations. These included the use of a database, geolocation, cognitive capabilities, a point-to-point topology that would allow users to connect only with a base station and not each other, and the requirement that the base station with which the devices communicate must be professionally installed with an outdoor antenna 10 meters in the clear.
MSTV said the record was insufficient for writing a rule. "It is difficult to see how repeated failure proves the concept," MSTV said. "Failure means the FCC does not know the sensing level that is necessary to protect TV viewers. As a result, the FCC will not know if devices built to any proposed 'sensing' rules will be sufficient to protect consumers. Moreover, after nearly four years, white spaces advocates have failed to build a device that works under controlled experimental conditions. There is absolutely no reason to believe that they will be able to build a device that will work in the real world."
