Incentive Auctions: A Plan for Success

One of the final questions concerns why TV channels 32-51 were chosen for conversion to broadband service. The answer is really a series of answers...

More broadband spectrum is needed for the commercial operators, but where will it come from? The FCC has pledged to “find” (meaning reallocate) 500 MHz of spectrum within ten years from when it presented its National Broadband Plan to Congress in the spring of 2010. Where will it come from, who will lose what they are using today, and how long will it take? One set of answers may be found in the reallocation of spectrum now being used by TV stations around the nation.

The FCC has been authorized to conduct “incentive auctions” in the existing TV spectrum. The incentive part of the auction means that any TV station that voluntarily gives up its 6 MHz in the upper portion of the TV spectrum will share in the proceeds from the sale and will be able to relocate lower in the TV band if it so desires.

I have not yet seen a proposed band plan to turn this TV spectrum into commercially viable broadband spectrum that can be auctioned and added to the spectrum already allocated to commercial network operators. There are many ways to divvy it up, but if this spectrum is to be used for Frequency Division Duplex (FDD), as is the rest of the allocated commercial wireless spectrum, the first determination must be how much separation there needs to be between the site transmit and mobile device transmit spectrum.

The original 800 MHz band, first allocated in 1981, had a separation of 45 MHz between the cell site and device spectrum (Figure 1). In 1996 when the PCS spectrum was allocated, the separation between the two segments was 80 MHz (Figure 2). In the AWS-1 spectrum, the separation is 100 MHz (Figure 3), and in the 700 MHz band, the most recent band made available for commercial broadband services, the separation between transmitters is 30 MHz (Figure 4).

Figure 1: Commercial Wireless Spectrum Currently Available in the United States

Figure 2: U.S. PCS Spectrum

Figure 3: AWS-1 Band and Potential AWS-2 Band (Source: FCC

Figure 4: Upper and Lower 700 MHz Spectrum the Upper D Block Now Allocated for Public Safety Broadband Service (Source: Phone Scoop)

TV spectrum is divided into 6-MHz blocks and for the sake of this exercise, I have chosen the spectrum used by TV channel 32 (578-584 MHz) and 51 (692-698 MHz). Today, channel 51 is the highest available TV channel and it is next to the 700 MHz broadband spectrum (known as the lower 700 band), “A” block that was TV channel 52.  This would provide a total of 120 MHz of new, prime commercial broadband spectrum. The issue is how to divide this spectrum so it has the highest monetary value per MHz for the auction, and so commercial operators can make the most efficient use of it.

LTE or Long-Term Evolution (4G) technology can be deployed in as little as 1.4 MHz (1.4X1.4 MHz) up to 20X20 MHz of spectrum in steps. However, the more spectrum available for each LTE system, the more efficient LTE is and the greater the capacity and data speed capabilities. The lower 700 and some of the upper 700 MHz band is divided into 6X6 blocks (one TV channel by one TV channel separated by 30 MHz, which is equivalent to five TV channels). For example, the lower A block is allocated 698-704 MHz and 728-734 MHz (old TV channels 51 and 57). In this spectrum, a commercial operator could build a 5X5 MHz LTE system, leaving spectrum on either side to further reduce the potential for interference.

However, LTE performs much better in larger band segments. The upper 700 MHz spectrum now has one 11X11 MHz assignment. Verizon owns the C block, which is 746-757 and 776-787 MHz or TV channels 60, 61 and 65, 66. The rest of TV channel 61 and 66 is set aside as a guard band between the C block and the spectrum allocated for Public Safety. The most recent allocation, 10X10 MHz of spectrum for Public Safety, is 756-768 and 788-798 MHz (all of TV channels 62, 67 part of 63, 68).

The “New” Spectrum

If we now turn our attention to the 120 MHz of TV spectrum from channels 32 to 51 as mentioned above and try out some different band plans, we can get a feel for how this spectrum can be best utilized. My assumptions are that the majority of the spectrum should be set up as 12X12 MHz with a mixture of nationwide and regional licenses. A few segments of 6X6 MHz could be used for more rural areas as well if the FCC would like to maximize the number of licenses offered. This works, and if we leave 6 MHz of spectrum vacant or for a guard band between the lower end of the new band and between the cell and device transmit bands, we could have four 12X12 MHz bands and one 6X6 MHz band (Figure 5).

Figure 5: One Idea for Reallocation of TV Channels 32 through 51 for Broadband Service


1)     A and A1 are guard bands, A is between broadband and TV channel 31, and A1 is between cell site and device transmit.

2)     This band plan provides four pairs of 12X12 MHz and one pair of 6X6 MHz. Further, the 6 MHz (F block) is adjacent to the lower 700 MHz A block and could be combined to provide another 11X11 MHz pair.

3)     Separation between cell site and device transmit of 30 MHz (same separation as both lower and upper 700 MHz spectrum today).

4)     Auctions could provide nationwide, MSAs, EAS, and REA if desired.

Of course, the band would be broken up differently. For example, instead of 12X12 channels, we could have 11X11 MHz blocks the same as Verizon’s C block in the upper 700 MHz band, and reduce the guard bands down from 6 to 5 or even 3 MHz each to gain at least one more 6X6 MHz segment. This could add as much as $4 Billion to the Federal treasury based on the Congressional Budget Office value of the 5X5 MHz upper 700 MHz D block at $2.8 Billion.

How many TV stations would have to opt into the auction or be relocated after the auction to make all of this spectrum available for commercial broadband use? Below is a chart listing the major TV stations licensed in the 32-51 TV band. This list was compiled using the FCC database and includes both high-power and low-power stations as well as TV translators. The approximate cost to relocate a TV station to lower in the TV band is estimated to be somewhere between $250K and $1 million depending on the TV station’s transmitted power rating. However, since the majority of viewers receive their TV via cable or satellite, in most cases they would not even notice that the stations had been relocated.



Total Stations


Total Stations










































Figure 6: Number of Licensed TV Stations in the United States on TV Channels 32 through 51

Note: TV station 37 is reserved for Radio Astronomy in the United States. This could be an issue in converting this channel into broadband spectrum. However, in the law authorizing incentive auctions there is already funding in the amount of $300,000,000 to move the users of TV channel 37.

Additionally, this chart does not take into account TV stations licensed in Canada and Mexico that might cause interference to broadband systems near our borders with these countries.

In total, 1,698 stations would be affected if all 120 MHz of the spectrum from TV channel 32 up to and including channel 51 were to be converted to commercial broadband spectrum. The issues to reallocate this spectrum, auction it, move the stations lower in the TV band, resolve the channel 37 Radio Astronomy issue, and deal with Mexico and Canada can be overcome, but this will be a time-consuming set of tasks. TV stations no longer require a full 6 MHz of spectrum per station because they are now digital, and some channel sharing is possible.

The auction process authorized by the Tax Relief Bill does not spell out how many TV channels should be part of the incentive auction. Further, there will really need to be two auctions. The first would be a reverse auction whereby the FCC will determine how many TV stations are willing to vacate their existing channel and relocate to a different portion of the TV spectrum or channel share with another station. Once this has been accomplished, there can then be a forward auction to determine what the spectrum is worth to the broadband community. If the numbers work out, a reshuffling of the TV spectrum will happen and some or all of the spectrum I have indicated above will be in play for commercial broadband use. The timeframe for the completion of both auctions is ten years, so none of this will happen quickly.

Ideal Broadband Spectrum

As can be seen from the beginning of this article, commercial network operators are already using spectrum in the 700 MHz, 800 MHz, 1900 MHz, and AWS-1 (1710-1755 MHz; 2110-2155 MHz) bands and the FCC is planning to auction the AWS-2 and 3 bands, adding another 30 or more MHz of spectrum in the 1700-2000 MHz band. Further, AT&T is considering using spectrum in the 2.5 GHz band, while Clearwire and Sprint are already using the 2.5 GHz band, and Dish network is asking the FCC for a waiver to build a commercial broadband network in the S-band (2000-2020 MHz; 2180-2200 MHz). If we add the 120 MHz of TV spectrum from channels 32-51 (578-698 MHz), it will make things really tough for device vendors. They will have to build specific devices for specific networks or a single device that covers 578 MHz through 800+ MHz, the 850 MHz, 1900 MHz, 1700-2100 MHz AWS bands, and perhaps the 2.5 GHz band. Our device engineers are very good at what they do, and software-defined radio technology will help. However, when you add GPS, Wi-Fi, Bluetooth, and Near-Field Communications (NFC) to these devices, the radio portion will become very crowded. If the device also needs to be capable of roaming internationally, the number of frequency banks increases by three or more. I have every confidence that there will be a single device at some point, but it won’t be easy.

What Spectrum Works for Broadband?

One of the final questions concerns why TV channels 32-51 were chosen for conversion to broadband service. The answer is really a series of answers; let’s start at the low end of the spectrum. Radio spectrum below 575 MHz or so is not well suited for devices. Yes, there is technology available today to build transmitters and receivers in all portions of the radio spectrum, but the other components have to be sized depending on the frequency band in which they are operating. Below about 575 MHz, the filters, duplexers, and antennas are too large and bulky for handheld devices. Antennas have to be tuned for a specific frequency and below 575 MHz they are simply too big. On the high end of the spectrum, today’s practical limit is about 2.5 GHz or 2500 MHz. Beyond that there are a number of problems. First, many more cell sites are needed to cover a given area than to cover that area at 600 MHz, 700 MHz, 800 MHz, and even 1900 MHz. Next, the higher the frequency, as a rule, the less penetration there is inside buildings. For all practical purposes then, the usable and most valuable spectrum for mobile or handheld broadband service is the spectrum between 575 MHz and 2500 MHz.

Within this swath of spectrum, there are currently other users including Public Safety, business and industrial users, satellite services, GPS, unlicensed Wi-Fi band, Bluetooth, and ZigBee, all in the 2.4 GHz band, and many military and government systems. It would be impossible with today’s technology to re-order portions of this spectrum to end up with an ideal scenario for commercial broadband.


The incentive auctions and TV band repacking authorized by the Tax Relief Bill enable the FCC to keep its promise in its 2010 National Broadband Plan. In this report, the FCC pledged to “find” 500 MHz of spectrum for commercial broadband services. If TV channels 32-51 can be repurposed for broadband services, this will add 120 MHz of ideal spectrum to that already available, and it will go a long way toward helping relieve congestion for wireless broadband services due to soaring demand.

In addition, this 120 MHz of spectrum could provide up to $40 billion in auction revenue. Some of that will go to the TV stations that elect to move during the incentive auctions and some will go to build the National Public Safety Broadband Network, but a large portion will go to the U.S. Treasury to help pay down our national debt, perhaps $25 billion or more when all is said and done. Spectrum is a finite resource so it won’t be easy, but this can be accomplished. If a sensible band plan is put into place, the TV industry will not suffer, commercial network operators or new entries will have this valuable spectrum available to them, and consumers and business customers will have relief from broadband network congestion.

The FCC has a lot of work ahead of it. Not all of the parties agree that this is the right spectrum to repurpose and there are likely to be some delays as groups petition Congress and the FCC to find the needed spectrum elsewhere. However, at the end of the day, this plan makes the most sense and hopefully it won’t take ten years for this spectrum to find its way into commercial service, and for the $billions to find their way into the U.S. Treasury!

Andrew Seybold


One Comment on “Incentive Auctions: A Plan for Success”

  1. […] Rest of the article here Click here to cancel reply. […]

You must be logged in to comment or reply.