This page is an archive from our previous website. Please check out our new website where you can read new COMMENTARY eNewsletters, TELL IT LIKE IT IS blog posts or Press Releases.
What I have to say to the FCC Commissioners and companies that might be lusting after this spectrum even if they have to provide 95% coverage within ten years is "good luck!"

95% Coverage and It's Free!

Tuesday, June 03, 2008

A few weeks ago, I wrote about Representative Anna Eshoo's proposal for a nationwide free wireless broadband network to be built out on some spectrum in the 2100-MHz band. This Representative, an advocate of net neutrality and open access, wants the network to be open and free, but she wants it to be porn-free.


Now Chairman Martin of the Federal Communications Commission has decided that the FCC should auction the 2155-2180 MHz slice of spectrum known as the AWS III band (Advanced Wireless Services III). Okay, there's more spectrum to auction and more money to pay down a little bit of the national debt.


But there are a few twists in this new auction. The first is that the winning bidder will have to cover 50% of the U.S. population within four years and, get this, 95% of the population by the end of the license term! Oh, and it gets better. Not only does the winning bidder have to pay for the spectrum, it has to build it out and then offer at least one tier of free-yes, free-broadband service to the Internet at speeds of 786 Kbps (no mention of uplink and downlink differences).


Sounds ideal for a Google bid to me. Full open access and free, meaning the network provider cannot charge money for it but I didn't see anything that says it cannot make you slog through a bunch of ads to get to what you want, which means that while it doesn't cost money for the user it is not free at all. I am having a hard time with all of this. Muni-Wi-Fi, which has taken a turn for the worst, Wi-Fi finally getting back to the last few hundred feet, Clearwire with its full access to the Internet, nationwide network, Representative Eshoo and her free Internet access wireless broadband plan and now the Chairman of the FCC saying that the winning bidder has to cover 95% of the U.S. population at 2100 MHz and that there has to be free service as part of the deal? Oh, I almost forgot, Google and others want to make use of the TV white space for additional unlicensed wireless broadband access.


I am not sure of the numbers here, but to cover 75% of the U.S. population at 2.5 GHz I do know it will take 65,000 cell sites, so I would guess that to cover 95% of the population at 2.1 GHz it will take more than 82,000 sites. This sounds like a typical radio system. For example, when I was designing two-way radio systems for public safety, we could achieve about 80% coverage of an area, for example, for about $10 million dollars, the next 10% for another $10 million and that left us with 10% of the population not covered. Today's estimates for the 700-MHz D block by people who know what they are talking about are if they have to cover 95% of the U.S. population on 700 MHz, the network will cost somewhere around $20 billion to build.


Do we need to solve our rural America lack of broadband issues? Yes, of course we do. Do we need another broadband operator in San Francisco where the population already has a choice of fifteen broadband providers including DSL and cable? Probably not. In Japan, when you get a license, you are required to cover the entire country-urban, suburban and rural. The assumption is that you will have enough customers in major metro areas to enable you to afford to build out the rural areas, but Japan is a lot smaller than the United States. I have to wonder what people are thinking when they come up with these ideas. Who is going to bid for spectrum, build out a $20-billion network in less than ten years and give away one tier of service for free? The only answer I have for that is Google or some other company that believes it can make more than $20 billion in advertising revenue from the network.


Perhaps Congress should require any FCC Commissioner to have worked in wireless or at least the telecommunications industry, not as an attorney, before being appointed, and perhaps someone should open up a spreadsheet and run the numbers. I know, several companies including M2Z Networks offered to build such a network if the FCC would give it the spectrum, but the FCC turned down the offer. M2Z's webpage talks about the upcoming FCC meeting to discuss and perhaps approve this new auction and discusses its belief that using geo-tagged ads (location-based) it can pay for this network.


I ran through the cost of building out the network, but not the cost of operating it. Let's do a little math and say that it will take about 82,000 sites to cover 95% of the U.S. population. And let's say, just for fun, that each site costs about $800 per month in site rental plus another $500 per month in backhaul, and perhaps another $100 in maintenance. We all know these figures are very low, but this means the monthly opex cost of the network is $1,400 times 82,000 sites or $114 million, which is $1.37  Billion a year. Think you can do it for less? What if the average site cost was double? We still need to include insurance, electricity, back-up power, taxes, a billing system, a back-end system to make it work and other details. This would take opex to over $2 billion a year.


Okay, Google and M2Z, here are some numbers for you to crunch. Cost of the spectrum: a guess of $3 billion, cost of the network to cover 95% of the U.S. population $20 billion, normal monthly operating costs $2 billion per year or $20 billion over ten years for a total of $43 billion. That is a lot of ads to sell, and if the network is not built to 99.999% reliability, even if 786 Kbps is free, who is going to tolerate all those ads except those who cannot afford to pay for Internet services. And I doubt if these users would be a market advertisers would really want to address.


What is it about this country? We want great, reliable communications services, high-speed Internet access, and yet we are not willing to pay for it! Well, I believe that those who live in rural America and do not have access to the same data speed as the rest of us do will gladly pay for the service-it is better than doing without or paying for a two-way satellite connection with all its latency.


Muni-Wi-Fi was supposed to be about providing broadband to those who could not afford it, to help them connect to the Internet and improve their lives. The problem was not only did these systems not work, those who could not afford to pay for service could not afford the computer to use the free service.


What I have to say to the FCC Commissioners and companies that might be lusting after this spectrum even if they have to provide 95% coverage within ten years is "good luck!" I don't believe this idea has been thought through carefully enough. I am amazed at the companies that have invested in M2Z-big venture companies that expect a big payday. Please tell me how you expect to realize a big payday when you will have to spend $43 billion simply to provide free Internet services and those you can charge for the service already have five or more options available to them?


These are just my thoughts. They won't matter because the FCC Commissioners will move ahead anyway and five years from now they will be able to sell the spectrum all over again!


Andrew M. Seybold


COMMENTS: This is an archived post. Commenting is no longer available.

Frank Bulk - 06/03/2008 23:08:20

What's more likely to happen is that a company will purchase the spectrum and seek extensions/exemptions from the requirements. Yes, they would build out to a few areas, but eventually the network would peter out.

If anything, this spectrum should be opened up for unlicensed use -- much more likely to get use and success there. Of course, that won't achieve our national broadband policy, but I don't think a spectrum auction should be the chief means to achieve it, anyways.


Frank Bulk - 06/04/2008 00:12:22

Please don't think that I'm proposing a free lunch, and I understand that shared access does mean that interference can degrade throughput and quality.

But there is already lots of licensed spectrum to provide reliable data services 3G at 850 and 1900, WCS at 2.1 GHz, and 2.5 GHz. But how much broadband wireless is deployed?

A new unlicensed band operating at a lower frequency is going to be more attractive than 2.4 or 5 GHz. And perhaps this band could be unlicensed, but restricted for MAN/WAN, not LAN deployments, so that those who do deploy perform some voluntary frequency coordination.

Joe Nordgaard - 06/04/2008 11:13:04


Thanks as always for providing a healthy dose of reality.

A couple more to through on the pile:
- What of the Katrina mandate? Got back-up power?
- E911 for VoIP?
- Must carry traffic for social mandates? (schools, libraries...)
- Got masts? The masts in my area a pretty loaded.
- Are we building to achieve free indoor coverage...65,000 BTS could prove light at 2.1 Ghz unless you can kick up the power. (During the 3G standards effort, the German Gov't filed a study that showed 8900 BTS for Germany's top 3 cities at this frequency. The US has many more cities of this size or greater. Ummmh)
- While on a give-away free-for-all, how about public service announcements mandates? (There appears no end to what pandering politicians are willing to give away...)
- And what about protection from lawsuits if you have spotty cover? Or outrages? Will you be protected as a common carrier without all the mandates? What of political favors for coverage??? (Just call your Congressmen for coverage after all what leverage would you have as an individual asking for free service?)

I recall Chile once used an negative bid system (ie where the Chilean Gov't paid you for monopoly coverage rights.) Someone at the FCC should do an NPV on $43B and see where a negative bid lands even with billions in ad revs factored in. (I can see trying to navigate with google maps with three ads...between you and your next turn. )

The list goes on - it gets so, so...muni-wifi-ish.

As far as unlicensed "high powered" services...sorry... to me this is just another feel good concept a tad short on some important issues. The problem with "unlicensed on steroids" - with all due respect Frank, tower height is needed to achieve coverage - with no license - how many providers can cover using the same spectrum? 5? 10? Without interference coordination what happens to the "commons." and what of output power/emissions in a free for all environment - there is an additive effect here. (Here is a regulatory/legal can of worms in the waiting.)

While muni-Wifi never matured long enough to see competition (little surprise here) - why cover 300 sq miles of Philly and not just enter with say 20 sq miles of high value coverage? Muni-Wifi never got to the point coverage cream-skimming ala what MCI did to LD? If you try to ration who can provide the the RF commons and if restricted, will we have universal mandates for coverage? Gee that sounds like an evolution path to a license to me.

And lets not forget ... Capital loves a license not a free for all. (BTW forget being porn policeman for the state (and all that could entail when little John hacks it and mommy sues) ...what of gamers? Streaming content? Got bandwidth? What of alcohol ads? Cigarettes? Got a pet(a) to a regulator on the use of that "free" spectrum. Or better yet - a pandering politician with a microphone.)

Gee I guess you could call me a skeptic. :-)

Andrew Seybold - 06/04/2008 11:31:30

Frank and Joe, interesting comments from different sides of the fence, I tend to side with Joe on this one for a number of reasons he mentioned and because--and I will go back to my 1998 statement: "There is not a single, terrestrial wireless data only service which has ever made money", Perhaps M2Z or Google thinks that going to an ad supported system will make all of the difference but $43-50 Billion is a lot of money, even to Google, to run an experiment like this and it would be an experiment. Further, this is supposed to be a "WiMAX" ssytem in 20 MHz of TDD spectrum which means that in congested population centers they are going to have a bandwidth problem if they are successfull, but I don't think we have to worry about WiMAX too much longer--stand by for my next blog post later today!
Thanks to both of you for joining in.

Joe Nordgaard - 06/04/2008 11:49:12

with an interesting exception...the cdma450 EVDO operator (then Eurotel Praha, now Telefonica) in the Czech republic. With around 70K subs in the first year, all paying $34 per month, the service broke even in 3 months. It is still a cash cow until this day - still all data.

They did leverage their GSM net for a cheaper than greenfield launch, but this would have just moved the goal post about 12-18 months since the country was covered almost border to border with around 300 EVDO BTSs (without the voice switching issues - this was very cheap to buy and build).

cdma450 is a great untold story throughout Africa and Asia and will really start to show as China Telecom gets aggressive with their net - which is in some 30 Provinces already. (Andy - why not get q to take us both on the road to see how creative operators have gotten with this technology. ;-)

Daniel Callahan - 06/04/2008 13:45:48

I agree that this is a silly idea. As someone who works at a WiMax company addressing "Digital Divide" issues, and as someone who lives in the sticks, I get to deal with these topics a lot, personally and professionally.

Some comments to add:

First, not only does the FCC Chairman not understand wireless economics, I'd venture to say he also doesn't understand the economics of ad-supported business models. As we've seen with Muni Wi-Fi and other initiatives, people are not willing to slog through ads in order to get access. People are willing to put up with ads in order to use a service they find valuable such as email, video sharing, and so on. But you have to have a tremendous amount of click-through's to offset a $43B investment!

Second, if we want to provide Internet access to everyone as a matter of national policy, here's a shorthand way of framing the problem: I pay about $80/month for satellite internet access, and the same for 3G data. So basically we need to subsidize up to that amount to provide everyone access (ignoring capital costs and the fact that even satellite doesn't get everywhere). But as skeptics have noted, politicians seem to be looking for the proverbial "free lunch" here.

Maybe Google has the resources and business model to do this, but my limited interactions with them on wireless access topics leads me to believe they don't want to be in the wireless service provider business.

Lastly, Andy you are right on in noting that the issue for providing "universal" Internet access isn't just getting a feed to everyone. We've seen the One Laptop Per Child people come up short in their goal of producing a $100 PC. Anyone who's helped their less computer-literate friends knows that the average PC (or Mac) OS is not something that can be tinkered with by the average Joe.

Access matters, but it isn't free and isn't sufficient.

Frank Bulk - 06/05/2008 22:18:01

My apologies if I appear to be oversimplifying the issues here. My initial point was that I'm not so sure we need to auction this spectrum in the 2100 MHz band, or attach some kind of public broadband policy to it. Unless it's a band that has some near-worldwide spectrum harmonization, I think we use more unlicensed broadband -- those frequencies seem to get a good amount of use. My day-job is even using the 5 GHz band to backhaul cellular traffic!

An ad-supported system is a non-starter in terms of paying for a national broadband network. I'm not going to qualify it, as others have already noted that it didn't even make it in metro/muni-WiFi arena.

In terms of a national broadband policy, I'm not sure setting the upper bar to be the price point of satellite service is the approach we ought to take. First of all, let me point out that it appears that wireless has become the politicians' de facto mode of communication. I'm not denying that it's very cost effective in rural areas, but it also sets a maximum in terms of bandwidth to the home. If we truly believe that Japan's 100 Mbps is where we need to be, isn't wireless just an interim solution? Perhaps fiber to the home is not plausible for 100% of the population. Perhaps the policy needs to be fiber to the first 90%, copper to the next 8%, and wireless to the last 2%.


Joe Nordgaard - 06/06/2008 06:38:50

Great post - completely agree. When the Congress was budgeting for the 2100 and 700 MHz auctions, a reporter at the Congressional Quarterly asked me questions about the spectrum.

First question was about auction receipts - which I estimated ~$15B for AWS (which closed at $13.7B) and between $20B and $30B for 700MHz depending on the rules (which closed at $19.6B). At the time of the interview, the reporter was shocked because no one anywhere had anything close to these estimates. (Both proved very close and as the reporter - now at the WSJ - said after the 700 MHz auction closed, I was by far closer than dozens of people he interviewed from the law firms and economic and policy think-tanks that litter the landscape in DC.)

The reason for the range on the 700 MHz estimate I explained that the real issue for one of public policy: Would the FCC use the spectrum to create a serious third path into the home via broadband wireless? This I argued would require explicit guidance from Congress to achieve since most if not all, of the remaining spectrum would have to have been allocated explicitly for that task and would require a certain amount of market exclusion to achieve real competition to the cable/twisted pair/fiber embedded base.

I told the reporter that I seriously doubted the FCC would make the rule changes needed for such a sweeping approach - which I said if they didn't would be unfortunate since high quality spectrum such as the 700 MHz band with sufficient bandwidth would not be made available for many years to come - if ever. The opportunity was lost and the consumer may have been the loser in the end.

It is not difficult to realize that the current info/entertainment duopoly to the home needs some measure of competitive alternative to help counter balance their market dominance as well as provide reasonable solutions to the millions beyond the reach of the last 5 - 10 miles. While downlink satellite entertainment does offer some alternatives without a full bundle of entertainment and data/voice, I believe this path will increasingly notch back further in importance as cable responses to fiber. A wireless data/voice bundle would have help fortify this segment's viability into the future. This would have reeked of meddling in the market if the FCC did get this explicit in its noodling (further adding to my belief it was a non-starter.) (Dish Network's play in the 700MHz E-block could prove interesting. They will need a TDD solution to pull it off and co-channel is high powered MediaFLO at 25Kw output will be an issue many markets.)

Naturally, I knew my suggestion would not win friends in the industry - but I also knew little old Joe sitting in his office in suburban NJ chatting with a reporter would not come close to shifting the tectonic forces that move around DC. (I have this problem about honest opinion - it can get you in trouble. He asked - I told him.)

Having said all of that - I have been working with mobile operators around the world to utilize their spare capacity in suburban and rural areas to provide wDSL via EVDO and HSDP. This is especially an interesting opportunity for operators in both developed and developing economies where the former state monopolies usually provide both poor and expensive services. Two of my clients had took this issue a step forward in LATAM just this week. It was a very exciting development - wDSL is coming. We will hear more on this in the coming months.

We are all very lucky to watch and participate in a profound revolution in real time. Enjoy the ride.

(Sorry if you received this several times - I used copy, delete, paste as my editor after posting.)

Frank Bulk - 06/06/2008 08:26:22

It's the FCC's role to regulate the telecommunications industry for the benefit to consumers (both residential and business), not promote a certain technology. As you well know, the FCC has been pushing the BPL agenda for some time, but that technology is not taking off.

The FCC needs to let the industry choose the technologies for providing services. Where there are roadblocks to using certain technologies, wireless or wired, the FCC should remove them.

If wireless is the most cost-effective means to providing voice or broadband in rural areas, that's fine, but any kind of subsidy that the FCC is considering in this area must be balanced against the knowledge that wireline services will always deliver higher speeds than wireless, and so if the FCC perceives (as it should) that someday there might be an application that requires greater bandwidth than what wireless can cost-effectively achieve, they need to make sure they don't hitch their wagon to wireless-centric agenda that may preclude the incentive to make necessary investments in wirelline technology today. Reverse auctions for USF are one example of short-sighted thinking.

Joe Nordgaard - 06/06/2008 10:41:25

I was not suggesting a reverse auction...only it drives the point about the NPV on Andy's point about the $43B cost. It was sarcasm on the notion that the business may only be viable if the FCC paid someone.

Similarly I had maintained during the 3G process in Europe, that even if the EU gave the spectrum away and mandated a coverage schedule, it still would have been difficult to make a business out of a technology that was clearly 5 years from market reality. Instead regulators helped herd the operators into a series of auctions which ended up costing the industry in excess of $200B. It took years for the industry to recover and 3G is still less than a homerun to this day.

To your point, 3G is a classic mandated technology - imposed by EU Bureaucrats for industrial policy reasons. Had the operators been allowed to introduce 3G in "2G" frequencies (which the European operators have in abundance from years of liberal spectrum handouts), some of the "do or die" premium would have been taken out of the equation and a less financially crippling more financially rational auction outcome may have occurred.

Andrew Seybold - 06/06/2008 11:45:56

Frank and Joe--great postings--I guess I am going to have to add both of you to the editorial staff of ASI!!!

Thanks, I am enjoying reading these posts.


Joe Nordgaard - 06/06/2008 11:51:08

I see there was a postponement - maybe someone in DC sent them your editorial. ;-)

Andrew Seybold - 06/06/2008 11:55:55

Joe--thanks for th econfidence in me but I learned a log time ago that nobody in DC listens to me--or to most folks who don't have a big check book to help specfic "causes", so it must have been for some other reason, of course the folks at M2Z might have learned how to build a spreadsheet model and plugged in some figures--something I wonder why the high visability investors in M2Z did not appear to do.

Frank Bulk - 06/06/2008 20:54:40

Looks like rural telcos aren't going to take this lying down. This was from a newsletter:
"This week, rural carriers marshalled their forces to oppose a reported FCC plan to combine AWS-2 and AWS-3 spectrum into a nationwide license and to require the licensee of such spectrum block to provide “free,” broadband service. The FCC was tentatively scheduled to vote on an order adopting the plan at its June 12 th meeting. The Rural Telecommunications Group, Inc. (RTG), thirty-six individual carriers filing as the Rural Broadband Group, and numerous individual carriers filed comments arguing that the plan, however well intentioned, would limit opportunities for small and rural companies, and would hamper deployment in rural areas by destabilizing the marketplace and deterring investment in rural broadband deployment. In response to vigorous opposition, the FCC has removed consideration of the plan from the agenda of its upcoming meeting. "

Andrew Seybold - 06/06/2008 22:05:31

Yes abd Clearwire and the other AWS spectrum winners should be too, not only does it not make finanical sense it also does not make sense from a market perspective, if you want to incent people to build out rural America they have to have a way to make money and ad based revenue for populations of less then 30 people per square mile does not compute--and some of these people are weathy and would gladly pay for he services, others are workers but they still understand the value and would pay a reasonable fee--for more on my thoughts on Rural America broadband see my comments, filed with the FCC on the Public Safety D Block:

Thanks agaih Frank, there are very few voices of reason around Washington DC I fear and those of us who are not there don't get listened to I am sure.

Joe Nordgaard - 06/07/2008 05:43:54

2100MHz doesn't cover too much unless you really drive the power.

As 700 MHz is rolled out, operators will find that covering rural America is easier and more affordable but this will take some time to achieve.

Andrew Seybold - 06/07/2008 10:22:08

Joe--my point exactly--it is about how many POPs can I cover with a given cell site--each site, back-haul etc costs money, and there is a number of subcribers that supports that cell--or you share the over all cost accross a network--however at 2100 MHz it is a LOT different than at 700 MHz, I am working very closely with the Rural Folks and this is why I have proposed to the FCC that we use the D block and first responder network to sever Rural America as well--and with the existing network operators taking part in the venture.

Joe Nordgaard - 06/07/2008 11:03:45

Thus my push to get 450... and how that is helping the world outside the US for all the same reasons.

African operators went out of their way to thank me when I presented in Nairobi a couple of years back. Many said they would never use copper again.

We are clearly in the same page...which is a great place to be - (aligned with "Seybold") on the side of common sense, technology and physics to solve some very basic problems the world has grappled with since before Pheidippides ran the first Marathon. (Thank god for Google...I never would have dug that name out of my attic.) Cheers - have a great weekend, Joe

Frank Bulk - 06/07/2008 11:05:18

I think we'll find that 700 MHz will provide the necessary capacity for rural areas outside of town, but 700 MHz won't provide the necessary capacity inside of town. I think Airspan is probably one of the strongest proponents of the idea of using higher-frequency spectrum in town, where distance is not as important, and use lower-frequency (900 or 700 MHz) to serve the few locations outside of town.

Inside of town, I believe that generally the duopoly will probably do a better job/price point than a WISP.

Joe Nordgaard - 06/07/2008 11:48:52

This was also part of my argument for entering the 450 band, in 1998-99 I argued that if the EU maintained the ban on using 3G in the 2G spectrum - it would force hyper speculation which would mean licenses would be coughed up in the after market for other technologies.

If a company could get hold of 450 and marry it to 2100 MHz - it would stand a competitive chance in Europe (albeit thin). (Thus the formation if Inquam Ltd. of which I was a founder)

The Russian 450 operators and China Telecom both turned and asked this very question once the cdma450 solution was put in front of them. CT interests in 450 may appear to have been put on the back burners shortly after this due pressure from Japan and the EU, but I think people will be amazed at what they roll out the coming months.

Recently, CT privately celebrated 100M subs on WLL and if you check they have not been investing in PAS/PHS in a long time. When visiting a CT CO in Chengdu last fall, there were some 50 burned out PAS BTS sitting on the floor of the lobby. Sichuan has nearly 80-90% coverage with cdma450. Rural China is getting very good 30 provinces.

And the cdma450 network was instrumental in helping relief efforts in last months quake. I am naturally proud of these efforts having been awarded a Gold Medal of Honor by Huawei (the first outside their country and company in their history) for helping them open China (and the world) to this technology. (But when you work in an office in your house(or on the road)...and don't get to tell many people...the result is I get to burden you. :-) )

John Hemphill - 06/23/2008 10:56:46

I always enjoy reading Andy's of the few level heads in the industry. I am always amazed to read news articles referring to folks who should be aware of the massive infrastructure investment involved in building and maintaining a nationwide network, but aren't.

The hysteria around WiMax is another good example. WiMax may be very successful based on its merits, but I constantly read articles (many from those who should be better informed) that WiMax will be the holy grail....coverage to even the most remote underground fallout shelter and it will all be for free!!! No, any network that takes $20 billion dollars to build and maintain is not going to be free. Until wireless broadband comes magically out of the ether, there will always be massive and expensive infrastructure needed to support it.