AWS-3 and TVWS AgainTuesday, September 02, 2008
So the FCC is heading off to Seattle to run some tests on the AWS-3 proposed band of 25 MHz it wants to auction requiring that 25% of the network provide free 786-Kbps wireless data services to 95% of the U.S. population within ten years. Meanwhile, the TV White Space unlicensed versus licensed spectrum battles are starting to heat up and I have to wonder if either side has taken a long hard look at this spectrum in some of the major cities.
T-Mobile is in the process of launching its 3G network in the AWS-1 spectrum around the country and has complained to the FCC that using AWS-3 spectrum for broadband services could cause interference to its 3G network. So the FCC engineers have packed up their gear to run tests and are heading to Seattle, home of T-Mobile and were it has its 3G services up and running. It will be interesting to see the results of these tests but they will not be proof positive if they show no interference—the T-Mobile 3G network is hardly loaded, and testing in a few locations instead of system wide will not really prove anything either. You might remember that Nextel and the FCC tested the Nextel system before it was permitted to start delivering cellular-like services in a band that was designed for two-way radio systems with high sites and high power. The test results showed no problems but we know now that many first responder agencies were (are) plagued by interference from Nextel’s low cell sites. The result has been a multi-year effort called “rebanding” where the entire band is being re-organized so first responder channels are located in one portion of the band and Nextel channels are located in another section, not intermingled as they originally were. This process has been long, tedious and very expensive for Nextel and to some degree for other organizations using this spectrum. And even though the process that was started years ago was supposed to be completed by this year, current indications are that it won’t be finished until 2010 or later.
One of the reasons for T-Mobile’s concern is that every radio puts out spurious emissions outside of its allotted spectrum but as long as the offending signals are below a certain level, they meet the FCC requirements and the system is certified as being in compliance. This is one reason TV and AM and FM bands are normally licensed only for one-way transmissions. Since these stations are operating at high power, each tends to “bleed” over into the next piece of licensed spectrum. Again, the amount of power permitted in the adjacent channels is limited by the FCC, but when mixing one-way and two-way systems, there is a pretty good chance that the one-way transmission will interfere with the two-way signals.
T-Mobile’s system is built like any other cellular system. Licenses are broken into two segments and ALL licensees in this spectrum are required to use 2110-2155 MHz for cell site transmissions and 1710-1755 for mobile devices. This ensures that even if there are some spurious emissions into adjacent channels, they will not cause interference to another system. As a further explanation, the cell sites transmit 24/7 with a lot of transmitter power and the wireless devices transmit only as needed with very low power. In order for the cell site to be able to receive the low-powered signal from the wireless device, the site must be relatively clean of interference at the receive frequency because the receivers are listening to very low-powered wireless devices.
The potential AWS-2/3 band on which the FCC wants to authorize nationwide broadband services would occupy the 2155-2180 MHz portion of the spectrum, just above the existing AWS-1 spectrum. But unlike the AWS-1 spectrum, both the cell site and the device transmitters would use the same portion of spectrum, using technology known as TDD or Time-Division Duplex. The historic cellular systems and AWS-1 spectrum split the cell transmitters and device transmitters into what is known as FDD or Frequency Division Duplex. There is nothing wrong with TDD technology. It works, and it is being used in the 2500-MHz band by Clearwire and others for WiMAX. However, as stated above, even if ALL of the systems meet ALL of the FCC requirements for out-of-band emissions and power, there will still be issues for handhelds on the T-Mobile or other AWS-1 networks that are near an AWS-2/3 cell site, just as there were with the Nextel/public safety sites. The receivers at cell sites are in a fixed environment, they have filtering and high-gain antennas and they can be optimized to minimize any interference that may be generated at a cell site.
However, the handsets, which will be at least dual-band in the case of T-Mobile (1900 PCS and AWS-1), do not have the same filtering or receive capabilities as their cell site cousins. The handset receiver is usually the weakest link in a cell system, and the same can be said for TV receivers in a home or office because these receivers have to be capable of receiving over a wide range of frequencies. T-Mobile does not have licenses for the same portion of AWS-1 spectrum in every city, therefore the receivers must be capable of “hearing” a 5-MHz portion of spectrum anywhere between 2210-2155 MHz. The AWS-2/3 band starts at 2155 MHz so you can see that there is real potential for interference to the T-Mobile handset and notebook devices from AWS-2/3 cell sites.
This is one reason, by the way, that the FCC has licensed slivers of spectrum called “guard bands” between some of the new 700-MHz channels to be used for lower-powered systems and to protect the public safety spectrum. When MediaFLO, Qualcomm’s high-powered digital mobile TV service that uses the old UHF channel 55 was first deployed, Qualcomm had to be concerned about interfering with TV stations on channels 54 and 56. Both of these channels have since been cleared, but in the early days of the deployment, this was a concern because of the same issues. It is now reported that the FCC will conduct what it considers to be “conclusive AWS Interference Tests” on AWS-2/3 and then make its own recommendations. These tests will probably be just as “conclusive” as those run prior to Nextel being given the go-ahead to convert two-way radio channels intermingled with others, including first responders channels, into a cellular system with low-level tower sites.
There is 55 MHz in what the FCC wants to turn into the AWS-3 band for nationwide broadband coverage. Perhaps it could hold back the lower 5 MHz as a guard band to help protect T-Mobile and other operators that have paid a lot of money for their AWS-1 spectrum. By the way, when I penciled the cost of a network at 2100 MHz that would service 95% of the U.S. population, my ten-year projection was that the capex and opex would run more than $40 billion. That is one heck of an expensive network to get free 786-Kbps broadband to those who cannot afford to pay for Internet access. Even Google would have a tough time raising that much money in advertising revenue.
TV White Space
The other battle that is brewing is over what is referred to as TV White Space (TVWS) and I have written several articles about this. The idea, again floated by companies that are Internet-centric and not what I would call wireless savvy, is to open up the space between TV stations in the United States to unlicensed broadband use. The opponents, the latest being Qualcomm, Verizon and the CTIA along with the NAB and others, are saying that the spectrum could be used but should be licensed, not unlicensed.
If you read my columns on a regular basis, you know I agree with this. If there is a license holder, then a single focal point can be identified to work on in the event of interference. Again, the issue is interference to TV receivers in homes and in mobile devices (if the TV stations start broadcasting directly to mobile units as some are planning to do). The main reason the FCC licensed TV channels that are not adjacent to each other is because many TV receivers would suffer inference if there were two channels next to each other in the same city. There are some natural breaks between channels already. For example channels 2, 3, 4 are adjacent to each other; channel 2 is 54-60 MHz, 3 is 60-66 MHz and 4 is 66-72 MHz. This is why you might find a channel 2 and 4 in use in the same city but never 2 and 3 or 3 and 4. (See TV Channel by Frequency Chart)
There is a gap between channel 4 and 5, which is where FM broadcast, aircraft communications, a ham radio band, business and first responder channels and some government two-way radio channels are located. The TV channels pick back up with channel 7 starting at 174 MHz and the next break is between channels 13 and 14. If you remember, before UHF TV, most TV sets stopped at channel 13, which goes to 216 MHz. Channel 14 starts at 470 MHz and between 13 and 14 there is a lot of government spectrum, more ham radio, more business radio, first responder systems and some wind shear radar and other services. There are no more gaps in frequencies until the end of the TV band, which used to go up to channel 69 but now ends with channel 51, or will after February 2009. So the highest channel available for TV will be channel 51 (668-674 MHz) and the channels now on the air above channel 51 will be relocated down into the band.
Today, each TV station has two assigned channels—one for analog TV and one for digital TV—but again, in February 2009, TV theoretically goes all digital and analog channels will be turned in to the FCC. In the existing white space in major TV markets, the spectrum is already being shared, first with business and first responder channels, then with low-powered TV stations (universities, etc.) and with wireless microphones. The FCC’s plan calls for all wireless mics to have “beacons” to announce they are on and operating so unlicensed users won’t step on them. This means every wireless mic in use today will have to be modified or replaced.
Looking at a 160-kilometer or 100-mile radius of Los Angeles City Hall, we find an amazing number of TV stations in service. In fact, in this 100-mile circle around LA there are only a few channels that do not have TV stations listed directly in LA proper, and that is because they are being used by LA City and LA County for first responder systems and other two-way radio systems (Channels 14, 16, 20).
After the TV stations above channel 52 are moved down to the lower channels and after the existing TV stations give up their analog channels (if they do), the LA Basin area will still have slim pickings for TV White Space. I have to believe you could run this same search in most major metro areas and find the same issues. If the FCC only authorized the use of white space where there were two clear channels (12 MHz) with, say, a Megahertz on either side for a guard band, leaving 10 MHz of bandwidth, which is about what is needed for decent broadband capacity and speeds, I don’t believe you will find much spectrum available in any major city. If you permit unlicensed operation in one channel of white space (6 MHz) and hold back one-half of a Megahertz on either side for a guard band, that would leave a 5-MHz swath of spectrum, which today’s WiMAX is not suited for, but I suppose you could fit a higher-powered version of Wi-Fi into it or revamp WiMAX to make it fit into it that space.
I don’t have a clue how much power those pushing for unlicensed spectrum think they want or need, but if they plan on serving a lot of homes with yet another broadband link, they will need a lot of sites to provide decent data speeds. At what cost? Free because of advertising? I still don’t see a business model here, even for licensed spectrum in areas where there are already many choices. Perhaps you could make a case for using Clearwire to backhaul and using something like Wi-Fi access points mounted on telephone poles around the city for white space usage. Penetration into buildings might be better, but where is the business model?
Those who subscribe to the “free Internet,” therefore free wireless access, don’t understand that Internet access is not free. Once in a while I might be able to connect to a free access point, but from my home or business, I pay for access and, at the moment, I pay a premium for mobility when I use my wide-area broadband. But my service stays up and the companies running it are making money, some of which is reinvested to build out in other areas to provide better coverage, and some of which goes to the stockholders who paid the bills for the networks in the beginning. Free market does not mean no-cost market, it means choices so people can make their own decisions. As history has proven too often in the wireless marketplace, free market means overbuilding and then mergers and acquisitions or failures until we are back to a point where those left standing can make a decent return on investment.
Both of these concepts—AWS-3 with 95% coverage and TV White Space—are being promoted as the next big thing in wireless broadband, better than what we have now, free service and great coverage. Just a few years ago, we were hearing the same thing from a group of companies that were busy convincing cities all over the nation to deploy Muni-Wi-Fi services. Where are those companies today? They may still be around, but they are no longer in the Muni-Wi-Fi business because of a flawed business model. I wonder how long this latest batch of companies will be around.
Andrew M. Seybold