Does Microsoft/Philips Testing Really Prove White Space Devices Won’t Interfere? - TvTechnology

Does Microsoft/Philips Testing Really Prove White Space Devices Won’t Interfere?

Last Thursday Edmond Thomas met with several people in the FCC’s Office of Engineering and Technology. He presented field tests results showing that the Microsoft/Philips sensing receiver detected DTV signals above -114 dBm 100 percent of the time and thus, he claimed, protected TV viewers from interference from unlicensed white space devices.
Publish date:

Last Thursday Edmond Thomas met with several people in the FCC’s Office of Engineering and Technology. He presented field tests results showing that the Microsoft/Philips sensing receiver detected DTV signals above -114 dBm 100 percent of the time and thus, he claimed, protected TV viewers from interference from unlicensed white space devices.

Readers may remember Edmond Thomas was appointed chief of the FCC’s Office of Engineering and Technology (OET) in January 2002. He is currently working as an engineer and partner in the law firm of Harris, Wiltshire and Grannis, LLP.

After reading the presentation, I’m not convinced that it shows the sensing circuits tested will actually protect DTV reception. I’ll be referring to pages in the presentation, so you may want to download Microsoft/Philips Field Tests and follow along.

Page 5 of the report notes that if the protected signal level is -84 dBm at the receiver, sensing at -114 dBm will provide a 30 dB margin. This is true only if the antennas used for the DTV receiver and the sensing receiver are the same, which, as I’ll show later, almost certainly won’t be the case. I should point out that some comments filed in the proceeding showed the sensing receiver would have to detect signals well below -114 dB to prevent interference.

Anyone that’s used an indoor TV antenna may be surprised at the findings on page 10. The report states “…no TV signal strength variation in excess of 20 dB was observed. At many test sites, signal strength variation was less than 10 dB. Within a room, TV signal strength variations typically did not exceed 5 dB.”

Based on my experiences in viewing DTV signals in hotel rooms across the country, moving the antenna a foot or two can make the difference between reception and no reception in many cases. The variation was certainly more than 5 dB. Could the choice of antennas used for the sensing receiver have affected the results?

Page 10 also notes that both devices must determine the channel is clear before communication starts. This should greatly reduce the chance of interference from a white space device transmitter that doesn’t see the TV transmitter. However, what happens when they are trying to connect to each other? If the devices transmit on frequencies they believe are clear, but which the other device does not see as clear, won’t that cause interference until a channel that’s clear at both ends is discovered and the units connect?

Going back to the discussion of the sensing antennas, one with a few dB of gain will detect a signal weaker than an antenna with no gain. An antenna that matches the polarity of the transmitted signal will detect a much weaker signal than one that is cross-polarized.

According to the FCC’s TV Engineering Database in CDBS, as of Sept. 19, 2007 none of the stations used for the Microsoft/Philips field tests transmitted any vertically polarized signal. The authorized facilities show horizontal polarization only. Page 9 of the Report states measurements were made using a monopole or discone antenna. Both of these antennas are vertically polarized. While they could be rotated 90 degrees (placed on their side) to achieve horizontal polarization, that would make their azimuth antenna pattern quite directional, with much lower gain in some directions. It has to be assumed the tests used the antennas in their normal position, which favors vertically polarized signals.

This is significant, as cross-polarization can reduce signal strength by 20 dB or more. Thus, the 30 dB margin assumed in the Report would in reality be much, much less. If the receive antenna has some gain (such as the Silver Sensor) and a preamp (the Terk HDTVa, for instance), the 30 dB margin would disappear completely. The choice of vertical polarization also explains why the variation in a room or building was a lot less than NAB and MSTV found (and what many readers have probably experienced). Because of the cross polarization, the vertically polarized signals likely arrived at the sensing antenna by way of reflection, as the signal bounced around the room. There may have been a significantly stronger horizontally polarized TV signal--the type detected by most indoor TV antennas--but the vertically polarized sensing antenna would have attenuated it.

Was using a cross-polarized sensing antenna simply an oversight? Or was it an attempt to make sensing technology (that broadcasters have said is not appropriate for this application) appear to work, when a properly conducted testing, such as that conducted by the FCC shows that it won’t? As a result, the Microsoft/Philips tests results cannot be used to show that sensing technology is a viable way to prevent interference to over the air DTV reception.

The Microsoft/Philips tests don’t address the issue of interference to cable TV reception. As previously reported, the FCC found white space devices operating near a TV set receiving cable signals could cause interference. Since cable companies go to great lengths to make sure there is no white space on their cable systems, it is unlikely any white space device could avoid causing interference to nearby cable-connected TV sets.

Even if you ignore the cross-polarized sensing antennas, neither the Report nor the FCC testing considered that with the adoption of new DTV transmission technology such as A-VSB or MPH, DTV could be received on portable and hand-held TV sets with signal-to-nose ratios as low as 4 dB. While this increases the immunity to interference, it also means the portable sets would be able to receive weaker signals, another indication that -114 dBm, even if used with correctly polarized antennas, is unlikely to be sufficient.

Unfortunately, recent articles indicate regulators and the general public don’t get it. The perception is that technology will somehow protect their TV reception and give them these wonderful unlicensed wireless devices. Of course, unless they are living in a huge mansion on a country estate, there won’t be any improvement in coverage over existing IEEE802.11-pre-n unlicensed devices operating at 2.4 or 5.8 GHz. TV broadcasters do not oppose licensed use of this white space spectrum, even if transmitter powers are much higher than those proposed for unlicensed devices. These licensed links are the ones that will truly deliver broadband to consumers and to rural areas, not low power unlicensed devices.

There are currently more than 10,000 comments in this proceeding. Most are one-page email messages asking the FCC to approve unlicensed devices in TV spectrum. I doubt that many of the commenters realize that allowing a free-for-all on these channels will not only interfere with their TV viewing on indoor, portable and hand-held TV sets, but also cause problems with cable TV reception. Blind faith that technology will solve the problem doesn’t guarantee that it will. TV broadcasters are limited in their ability to adapt technology to new environments, as they can’t disenfranchise all the viewers with older TV sets. The FCC wouldn’t allow it.

In Wednesday’s workshop on the DTV conversion, some commissioners commented about the possible reaction from viewers if they should lose TV on Feb. 18, 2009. The comments are available at If that concerns them, what about the potential for reception problems when unlicensed devices in the band are turned on? Consumers purchasing DTV converters, new TV sets, or a cable subscription giving them another three years of analog TV, may start seeing their TV pictures go to blocks or freeze (if digital), or to snow (if analog), a few months after the transition if unlicensed devices are turned on.

The public will remember the government made them change their analog TV, and it’s safe to assume they will blame the government and the FCC for making them switch to a system that doesn’t give them reliable reception. They may not connect the interference to white space devices, but they will remember the old system worked and the new one doesn’t. They won’t blame Microsoft and they won’t blame Philips, or Edmond Thomas’ law firm, but they will blame the FCC and will likely complain to their local TV or cable company and their Senator or Congressman. Any decision on white space devices has to be based on solid engineering along with plenty of safety margin.