The Spectrum Sandbox and GoogleTuesday, May 29, 2007
On May 21, Google sent an Ex Parte Filing to the FCC regarding the 700-MHz spectrum. In my Blog, TELL IT LIKE IT IS, on May 24, I mentioned that Google has revealed its interest in bidding on 700-MHz spectrum at the upcoming auctions.
Not wasting any time, on May the 24, the FCC issued DA 07-2197 requesting comments on the proposals Google made in its letter, and asked for these comments to be provided within seven days after publication in the Federal Register with reply comments due no later than fourteen days after publication.
I have read both the Google Ex Parte filing and the FCC's request for comments and have some comments about this matter based on what I have been able to extract from the two documents and from talking to various people familiar with other 700-MHz proposals. I will be filing comments in this matter because, while Google's proposal may be well intentioned, I believe that neither the technologies nor structure proposed is the best solution to the spectrum scarcity issue.
Google requested in its letter that the FCC clarify that it already allows the use of dynamic auction techniques such as real-time auctions and per-device registration fees. By doing this, Google states, it will permit ubiquitous wireless broadband Internet access to all Americans, which is a worthy goal. Google also states that making it more practical for new entrants to bid effectively and successfully in the upcoming 700-MHz spectrum auction will bring innovative new broadband-based applications, services and devices. It claims that this approach will help accelerate the penetration and uptake of broadband services for consumers and bridge the so-called "digital divide" that continues to separate far too many Americans from the technological tools critical to economic, social and personal advancements.
The essence of Google's proposal, from what I can gather, is for spectrum to be made available on an
as-needed basis, that there be an Internet clearinghouse for making this spectrum available and that the ultimate goal is not to build networks as we know them today, but to provide a wireless connection to the closest Internet access point and let the Internet move traffic to where it needs to go, basically replacing the network with the Internet.
Google claims that most of the
Google goes on to say that The Commission should require broadband platforms in the lower 700-MHz band and calls out, specifically, the orphaned 6 MHz of spectrum known as channel 56, pointing out that this spectrum, currently designated as the E Block (722-728 MHz) be used only for 1) interactive, two-way broadband services, 2) to be connected to the public Internet and 3) used to support innovative software-base applications, services and devices.
One channel below the E block is the new Qualcomm MediaFLO, one-way digital TV system that operates on channel 55. This system uses high-power transmitters, and while required to meet FCC criteria for spurious emissions and other things, I believe that the E band, in many parts of the nation, will not be a nice, clean band for use by low-powered devices. Anyone who lives near an XM Satellite terrestrial transmitter knows that Wi-Fi within a radius of one of these transmitters is useless even though XM is meeting all of the required rules and regulations.
There is a reason AM, FM, and TV stations were grouped in the same bands. They transmit at high power and there is interference to adjacent channels. Regardless of what the software-defined community has to say, or the White Spaces Coalition has demonstrated in the lab, today's real-world environment is one of noise and interference that cannot be ignored.
I believe there is room for compromise as we look at our spectrum as it is used today and as technology continues to advance. But I have concerns surrounding many aspects of this proposal. The first is that Google wants some of the spectrum to be used on a secondary basis when it does not interfere with the main provider of services. It cites some technologies that are embedded into Wi-Fi and "prevent" interference. My own experience from testing Wi-Fi systems is that these technologies work only to a point. When the noise level is too high and there are a number of competing signals, the overall performance of any network is degraded.
Another issue I have is that Google seems to believe that the Internet is a mission-critical, hardened network. I disagree. The Internet is not an administered network with a common point at which all of its nodes and connections can be monitored. It is, to some extent, self-healing, but it is not a mission-critical network.
I do not take issue with the premise that we need to provide broadband access for every person in this nation, but I do take issue with some of this coverage being on a secondary basis. Even Google, the company providing free Wi-Fi access to the Internet in
If Google's statement that we only use our spectrum 5% of the time and it sits wasted and unused is correct, then the federal government should release more of its spectrum. How about the radio channels used by the Secret Service for the protection of the President and others? That spectrum sits fallow most of the time in 99% of the nation. Shouldn't it be included in the spectrum pool?
Perhaps we don't always use our spectrum as efficiently as we could. But Google's claim, also made by others, about how little spectrum we use on a daily basis is based on the fact that at any given point in time, spectrum for two-way communications is not being used. AM, FM and TV stations that are on the air 24/7 use their spectrum allocation 100% of the time. The National Weather Service broadcasts on 162.55 MHz (and a few other channels) 24/7 and our cellular systems are on the air 24/7, not only when someone needs to make a call or send or receive data, but all of the time.
The only time two-way radio communications systems are in use is when there is radio traffic to be sent or received and the spectrum must be available when it is needed. Perhaps the Internet community could better understand this if we use a dial-up access to the Internet metaphor. Discounting the slow speed of dial-up, people who have a dial-up connection do not dial up that connection until they want to go online, and then they want that connection. If the circuits are busy, they have to wait and redial to make a connection.
People with DSL, cable or wireless broadband access do not have to wait to make a connection, they are on the Internet 24/7. But in the case of broadband wireless and Wi-Fi, their connection speed will vary depending on a number of factors including how many people near them are also on the network.
I have read study after study about underutilized spectrum, and they are always performed during the day (I guess those conducting the study don't get paid overtime), they are usually made in good weather and, therefore, at the lowest usage point. If studies were conducted at 10 pm on a Friday or Saturday night (for first responders), results would show channel usage to be close to 100%. The reason communications are not impaired is that a dispatch center is in control of the radio channels and dictates who can talk when. A human determines the urgency of the communications and sets up priorities for channel usage.
One issue I have discussed before when it comes to sharing first responder and commercial resources on the same spectrum is that usage for both sectors peaks at the same time. When there is an emergency or a disaster, many people want to reach out to make sure their families and friends are alright while first responders are trying to communicate to bring the situation under control.
I believe Google made some interesting points and should be working with people in the wireless industry who understand interference and other not-so-apparent laws of physics that surround wireless. Wireless only works well when there is a gatekeeper-someone to manage the network and the flow of traffic-whether a dispatch center for first responders or a more complex network complete with a network operations center to balance the load and demand.
The Internet does not have a gatekeeper, but each company that uses the Internet for access does. Google manages its customers' demand for service. As it grew in size and scale, it added more servers and made other changes to accommodate demand. The idea of access to a wireless service without any type of gatekeeper is what made the Citizens band useless and it is what is keeping Wi-Fi systems from being reliable and trustworthy networks.
Perhaps some good will come out of Google's proposal, and I suggest that Google Internet architects sit down with wireless network architects and try to find a common language so they can gain an understanding of the each other's issues. I also believe that promising but untested technologies should not simply be authorized for use. Rather, a portion of spectrum should be allocated for them to be further developed and tested in the real world.
It seems everyone wants to play in the wireless sandbox. Those waiting their turn seem to think that those already there are bullies trying to keep them out. We cannot build a bigger sandbox, but perhaps we can find a way for those who want to play to climb in and out as they need access. In the meantime, we need to look after those already there that have invested huge sums of money to ensure everything works on a nearly-perfect basis.
The 700-MHz spectrum is too valuable to waste, but it is also too valuable to be turned into a sandbox littered with trash and garbage. Once again, I say we only have one chance to get it right-for everyone.
There will be other proposals coming to light, of that I am sure. One might even provide a better way to permit high-speed access to the Internet for everyone in the
Andrew M. Seybold