White Spaces Are There For a Reason

I have recently written about the hot topic of sharing broadcast spectrum with unlicensed devices.

There are those who believe that the television broadcast spectrum is being under-utilized, since there are channels with nothing on them in each locality. They call these "white spaces," like an empty piece of paper.

Moreover, they believe that new wireless services could make use of these white spaces.
Table of Primary Allotments of DTV Channels for Portland, Ore. and VicinityLow VHF Band Channels
2-6 Vacant, might be deemed white

High VHF Band Channels
7 KOAC Corvallis
8 KGW, Portland
9 Vacant, but both first adjacent channels are allotted locally
10 KOAP, Portland
11 Vacant, but both first adjacent channels are allotted locally
12 KPTV, Portland
13 Vacant, but is first adjacent channel to 12

UHF Band Primary and Secondary Allotments
14 K14HN
15 Vacant
16 KORS (Class A) will probably get this channel
17 Vacant but 16 and 18 are allocated locally
18 KOXI (Class A) will probably get this channel
19 Vacant but both 18 and 20 are allocated locally
20 K20EH
21 Vacant, but first adjacent to 20 and 22.
22 KPXT, Salem
23 Vacant but both first adjacent channels are allocated locally
24 KNMT, Portland
25 Vacant but first adjacent to 24
26 K26EJ
27 Educational reservation
28 Vacant but 27 is allocated to Portland
29-31 Vacant, may be deemed white
32 Vacant but 33 is allocated locally
33 KWBP, Salem
34 Vacant but both 33 and 35 are allocated locally
35 KORK (Class A) will probably get this channel
36 Vacant, but 35 is allocated locally
37 Reserved for Radio Astronomy by International Agreement
38 KKEI (Class A) will probably get this channel
39 Vacant but both 38 & 40 are allocated locally
40 KOIN, Portland
41 Vacant, but 40 is allocated locally
42 Vacant, but 43 is allocated locally
43 KATU, Portland
44 Vacant, but 43 is allocated locally
45 Vacant, may be deemed white
46 Vacant but 47 is allocated locally
47 KPOU will probably get this channel
48 KPDX, Portland
49 Vacant, but 48 is allocated locally
50 Vacant, but 51 is allocated locally
51 KOXO (Class A) will probably get this channel

(Class A) denotes a Class A station. These enjoy protection against interference the same as that provided for full power stations. They will get DTV channels after Round 3. LPTV and translator stations will get DTV channels after Round 3 as they are secondary in priority. The FCC has not set dates by which LPTV & Translators must cease analog signal transmission. Out-of-Core broadcasters must find an in-core channel before the end of the third round of channel selections.
I've always wondered about the number of these white spaces, or channels, and in how many communities will they be available.

In mid-June, written testimony submitted to the Senate Commerce Committee gave me new insight.

"Even after the digital television transition ends early in 2009, every one of the 210 TV markets will have unassigned and vacant channels reserved for broadcasting, but not being used," testimony from the New American Foundation reads. "Many markets will have dozens of open channels. Vacant TV channels are perfectly suited for WiFi and other unlicensed wireless Internet services."

This made me curious about the number of such channels after the Feb. 17, 2009 analog shutdown. I downloaded "Measuring TV White Space Available for Unlicensed Wireless Broadband," from the New America Foundation Web site at www.spectrumpolicy.org , dated Jan. 5, 2006.

This paper analyzes 22 TV markets for the number of white spaces they will have after 2-11-09. Having just moved to within line-of-sight of the Portland, Ore. DTV towers, I looked at the data provided in this document. After 2-11-09, the paper says there will be 10 high-power TV channels, five Class A TV channels, five low-power TV channels and one "other."

These 21 TV facilities occupy 126 MHz of spectrum. The broadcast spectrum at that time will be Channels 2-36 and Channels 38-51, or 49 channels occupying 294 MHz. So, it is argued only 126 MHz of this spectrum is being used. The implication is that the remainder is white space.


I wondered how 1 watt transmitters can operate in first adjacent channels to any of these 21 transmitters whose ERP can be up to 1 million watts (60 dB above 1 watt). This column and several professional papers have noted that the power in first adjacent channels is 44.5 dB below the power radiated within the authorized channel.

Therefore the sideband splatter into each first adjacent channel is up to 60 dBW - 44.5 dB = 15.5 dBw.

That is a lot of interference in the channel of a 1 watt transmitter to overcome. The proposed ERP for these unlicensed 1 watt transmitters is 2.4 watts, so the noise in first adjacent channels is much stronger than the signal being radiated by these unlicensed transmitters. For reception, the co-channel noise must be less, not more than the signal. Furthermore, broadcasters employ very tall towers, typically 1,200 feet, which gives them the ability to transmit strong signals over large distances.

The IEEE has a Working Group 802.22 devising a protocol for unlicensed transmitters that may operate within broadcast bands. The chairman, Carl Stevenson, has written to me that the use of first adjacent channels is "off the table." I note that MSTV and Dr. Oded Bendov and I all agree that first adjacent channels must not be used by unlicensed transmitters.

As my readers may already know, I am also concerned that strong DTV signals on first adjacent channels may generate third-order intermodulation products, which appear as noise in these first adjacent channels. I am also deeply concerned that large number of DTV receiving appliances will be sold with tuners having little if any RF selectivity to keep signals on taboo channels from jamming DTV reception.

It seems that the testimony cited implied that first adjacent channels could be used. I strongly disagree.

As an example, I considered my own backyard. I used FCC channel selection data to construct a Table of Primary Allotments for full power TV and Class A TV stations in the Portland Ore. area.

All five low VHF channels might be usable by unlicensed transmitters. None of the high VHF channels would be usable by unlicensed transmitters.

The number of vacant channels that are not first adjacent channels is nine. Some of these may be used for displacement purposes or for translators and low-power allotments.

Existing DTV channel allotments on first adjacent channels are or may have to be co-sited to avoid harmful interference. Unlicensed transmitters cannot be co-sited; they will be randomly sited throughout the community, slowly at first, but eventually they might be commonplace.

If these wireless services involving unlicensed transmitters are successful, there will be a large number of base stations. These will operate at the maximum authorized power and radiate from higher elevations than rooftop consumer's antennas.

The New America Foundation testimony indicates 20 white channels in the Portland market. Using the FCC DTV tentative channel designations for the first and second rounds, which include only full-power TV stations, 10 stations are listed.

However there are also five Class A stations in this market, which will be allotted a DTV channel and at least three LPTV/translators for which DTV channel assignments have yet to be made. In addition, there are seven LPTV/translators with out-of-core channels. These have a priority.

As the use of first adjacent channels as white channels is now known to be impractical, very few of the previously reported white channels are viable.

I believe the most efficient use of spectrum by white channels would be in a band of their own, with a suitable guard band separating it from broadcast spectrum.

Perhaps the best place for these unlicensed transmitters is within the low VHF band, where there are very few allotments (43) found in the Table of Tentative DTV Allotments. These allotments (in parenthesis) are Channels 2 (12), 3 (five), 4 (four), 5 (15), and 6 (seven).

A simple low-pass filter at the output of each unlicensed transmitter in the low VHF band would eliminate interference to TV, FM and aviation. Channels 2, 5 and perhaps the lower portion of 6 would be white space for unlicensed transmitters. The upper portion of Channel 6 would serve as the transition bandwidth of the low-pass filter needed to protect FM, aviation and military spectrum users.

This buffer band may be significantly less than 6 MHz. The FCC rules presently proposed for unlicensed transmitters in broadcast bands would allow a transmitter power output of 1 watt in Channels 4-51. If 1 watt is sensible in the UHF band, it makes no sense in the VHF bands, especially the low VHF band.

There it would scale to 10 milliwatts based on the dipole factors for these bands. This is made clear in OET Bulletin No. 69. But, why not leave the proposed power level alone? Let the transmitter power output level remain 1 watt in the new 54-85 MHz wireless band.

Possibly some spectrum immediately below 54 MHz could be made available. The fact is that at low VHF, tuner-generated noise is not the limiting factor, manmade noise is.

Furthermore, if this wireless service is for sparsely populated areas, the people living out there have the space to erect tall Yagi-type receiving antennas to capture low VHF signals, much as they already do for TV reception.

Clearly, the FCC might have to decide the priorities between broadcast and nonbroadcast uses of the public's RF spectrum if the planning of this new wireless service is not done expertly. Complete testing is clearly necessary because we do not know the performance characteristics of DTV receivers being marketed, or about to enter the market.

Realistic specifications must be adopted for the spectrum of these unlicensed transmitters. Field testing will be required to prove that due diligence in testing has been accomplished.

In particular, translators provide millions their only TV signals. Will it come down to whether TV is to be or not to be? Perhaps the only practical compromise is for Channels 2-5-plus to be used for new wireless communications exclusively.

This sounds like the familiar broadband-over-power-lines notion, but without the power lines. Whatever happened to BPL anyway? As always, your comments are most welcome.