ISPs Losing Interest In Citywide Wireless Coverage

The New York Times is running a story about how hope is fading for the implementation of municipal wireless access in cities across the US. Major cities and small towns alike are finding that ISPs are withdrawing from such plans due to the low profitability of ventures that are similar to Philadelphia's incomplete network. We've previously discussed Chicago's and San Francisco's wireless status, and also some of the stumbling blocks other cities have faced. From the Times: "In Tempe, Ariz., and Portland, Ore., for example, hundreds of subscribers have found themselves suddenly without service as providers have cut their losses and either abandoned their networks or stopped expanding capacity. EarthLink announced on Feb. 7 that 'the operations of the municipal Wi-Fi assets were no longer consistent with the company's strategic direction.' Philadelphia officials say they are not sure when or if the promised network will now be completed."

Not really surprising

WiFi just isn't a good technology for ubiquitous Internet, which is what they want to be providing. It's fine for covering a coffee shop, but for this kind of scheme to be cost-effective they need to cover at least a city block with a single access point.

They're facing competition at both ends. They can't sell the service as 'Internet access in that place where you really want it' because often 'that place' already has free WiFi. They can't sell it as 'Internet access everywhere' because they don't have the coverage and their competitors, the mobile phone companies, do. Always-available Internet via my mobile phone costs about the same, per month, as via my cable modem (albeit with slower speeds and much smaller caps). For people who are willing to pay for Internet to be available all the time, that's a much better option than WiFi.

Re:

Very true. City-wide WiFi makes very little sense given how US cell providers are trying quite hard to roll-out 3G/3.5G/4G services. Right now, those are a little on the expensive side, but competition and better technology will eventually drive them lowe

Re:

I agree with you about the prospects for ubiquitous WiFi now, but I don't agree with you that WiFi just isn't the right technology.

With time and experience, the coverage problems were resolved. And the cost of rolling out these networks was quite low - ti

Re:Not really surprising

With time and experience, the coverage problems were resolved. And the cost of rolling out these networks was quite low - tiny compared to the costs of rolling out the 3G networks

Coverage is an easy problem to solve. You just need a lot of access points. A really, really, large number, in fact. Walking around campus, I see access points in almost every room. Looking into the distance, I can see the cell tower that gives coverage to the campus and most of the city. Individual access points might be cheaper than cell towers, but the amount needed to cover a given area aren't when you factor in the cost of wiring them all up and the cost of sending someone out to fix them when they are damanged (much easier to secure and diagnose faults with fewer towers). Since WiFi uses unregulated spectrum, you also have to put up with the fact that microwaves, private access points, and so on can all interfere with your network. If someone starts jamming your UMTS tower then you can get them arrested.

Profits

They wont be able to screw the consumer hard enough with what is basically free and open. ISP's want that pay per byte contract and are drooling for when they can bring it back.

Re:Profits

That's exactly it. Capitalism hates plenty, and it will destroy it when it finds it.

Re:Profits

Parent was modded as funny, but its the one of the main reasons Tesla's wireless power transmission never got serious funding and interest by investors.

From Wikipedia [wikipedia.org]

However, in 1903, when the tower structure was near completion, it was still not yet functional due to last-minute design changes. Tesla intended for the tower to demonstrate how the ionosphere could be used to provide free electricity to everyone without the need for power lines. Morgan, who could not foresee any financial gain from providing free electricity to everyone, balked. Construction costs eventually exceeded the money provided by Morgan, and additional financiers were reluctant to come forward.

And the fact that WWI broke out didn't help either...

Even if he could get it working on a mass scale, there was no way to tell if people were using it for free or how much they were using. I suppose in retrospect they could have put a meter on the other side of the customer power receiver, but the way Tesla envision is that the devices would be independent so putting a meter on each and every light bulb would be impractical.

With wireless, they can at least track mac addresses and force a radius login, but it still has similar problems so its easy to see why businesses are reluctant to move forward with it.

Re:Profits

Free and open? "Open" I can understand, but "free"? Where did you get that idea? Even when I was in school I paid for access via tuition. Government supported access isn't "free" either, you just don't pay directly. Nothing is free.

Although I'm no fan of corporate greed somebody, somewhere has to pay for the service.

Re:

Free means unrestricted, not "brought to you by magic faeries."

The opposite of free isn't "expensive", it's "controlled".

Money has no meaning beyond self-aggrandizement when there's enough to meet everyones needs.

It may not be commercially viable, but communally

If they think there isn't money to be made (and can't figure out the potential of giving cable/DSL subscribers free WLAN access on the road as an extra, much like Fon [fon.com] does), well, then, as has been proposed years ago, just let someone else do the job, such

Just a few more Linksys users

If there were just a couple more Linksys users in every neighborhood, we'd have city-wide access in every city in America! From my house right now, I "see" eight wireless signals. The three that are unsecured are all labeled "Linksys".

I still believe

I still believe that city-wide WiFi is an achievable and useful thing, but only if it is provided by the people for the people. The amount of work and the cost for any individual who allows the public to connect to the internet through an already existing wireless access point is very low and the benefit of being able to use other people's access points for free is high. Politicians should not seek to fund commercial WiFi deployments. They should provide legal protection to people who share their network connection with the public.

wow

EarthLink announced on Feb. 7 that 'the operations of the municipal Wi-Fi assets were no longer consistent with the company's strategic direction.'

Wow, EarthLink is still in business??

Re:

Wow, EarthLink is still in business??

Yep. They are. They're still offering dialup at $21.95 a month (with various ways to get that lower through comittments or bundling with other services.) They also have a very interesting $49.95 a month offer in select

We all have to sleep in it...

The ISPs made their own bed but we all have to sleep in it now.

The ISPs fought tooth and nail against even modest municipal wifi limited to public areas like libraries and shopping districts, because they wanted to make money from it. So rather than municipally funded projects they promoted these ad-hoc "partnerships" that didn't, in the end, make money.

They're too obsessed with $ now!

That's why the United States is behind many other major industrialized countries in Internet speeds. Because companies in the United States care about the $$$, not about innovation or advancing technology.

I guess the ISPs decided trying to wiretap MANs wou

Re:They're too obsessed with $ now!

Because companies in the United States care about the $$$, not about innovation or advancing technology.

While countries elsewhere in the world are altruistic, caring little about money?

Mzone

I just signed on with NTT DoCoMo for about $8 a month to get Mopera/Mzone which is basically 11Mbps 802.11 access points at places like KFC around town (Tokyo). There definitely aren't enough but maybe one near most major stations. However they require a lousy browser-based login (works automatically with their utility though) so I can't use my Skype phone.

It is very cool, and still only about $15/mo. even if you aren't already a customer of their FOMA phones.

Free is good but also maybe a very low monthly fee ($5/mo.?) to a general fund that would be divided among routers/isps doing this? Whatever, the question is find a way to get it done. I never thought ISPs would make much money out of this myself but once you get onto 11Mbps you get addicted. I use the windows app they give you to look at a detailed map of the city and find where the nearest point is. So far Tully's and Kentucky are my faves!

Why didn't they all collectively buy spectrum?

There was a perfect opportunity to provide wireless access for everyone... the Fed just auctioned off the very spectrum needed to make it a reality. What happened?

Did anyone consider this? I know Google had mentioned it and it was a meme floating around that they might buy up the spectrum and offer *free* wifi everywhere, supported by ads of course... if the States or a collective of cities had gotten together and purchased the bandwidth, it really could have been free.

Maybe I'm missing something? Was it not a great opportunity from a technical POV? or did all our local governments just drop the ball?

802.11b,g = Silly, WimAx could work briefly.

Trying to create a network by installing hotspots every 150 feet or even every 1000 feet or so if you extend the transmission range over public spectrum where any Joe may interefere with it with their own cordless phone, microwave and/or home WiFi router was a silly endeavour.

Now doing the same thing with public spectrum WiMax or UltraWideBand could work for 2-4 years until the next wireless technology improves upon it. It might work because it can cover much greater distances, so less antennas, and better ones (MiMo) and most importantly, tower equipment is expensive enough (though not very), that the average home user isn't going to be able to afford a transmitter to compete with it.

There is one gotcha, and that is that sharing some 50 Mbits over 3-20 KM would never work because most Metro cities are too densely populated. Unless you use deep packet inspection at endpoints and allow only SMTP and true HTTP/s traffic and deny all else. South Korea has standardized on WiMax for wireless, not WIFi, and their network is mostly complete.

Ultimately, there is a reason why Telco's pay Billions for private spectrum, because there will be (in theory) zero interference, and zero competition - although Google changed the latter slightly, lately.

PS. If you want to learn way more on WiMax read the wikipedia page, it is very informative.

Not ready for prime time..

As someone who recently moved back to Mountain View (Google-ville) and used Google WiFi initially, my experience is that it is not ready to replace wired ISPs.

Indoor access via laptop is problematic, for all the normal coverage reasons. The Google router is right across the street on a lightpost, but it is tough to get a strong enough signal indoors. A wireless bridge place in the window facing their router solved the signal strength issues.

With a strong signal, it worked well much of the day. Not blazingly fast, but fast enough to be useful (~1Mbps). But, during high usage times, 7PM-10PM, it became unusable. Packet loss was terrible, so performance was too, and many sessions timed out completely.

And this is with a very mixed environment, most people I have talked to used Cable or DSL internet access. So, Google WiFi wasn't supporting the whole street. But, even with the subset of users, it was too much for their service to handle. I quickly decided to go for cable internet, and this is definitely the right choice for anyone needing Internet for business / critical usage.

Maybe it could work with other improvements, I don't know enough about their infrastructure to say for sure. Some thoughts:

- Uplink issues: I think Google routers hub back into their network via wireless links. Maybe that is the piece that is not holding up at peak times. If so, a better network back to the ISP may help.

- 802.11N may help: More bandwidth, longer range. Operating on the 5GHz band may also be less crowded with other networks, at least for now.

Citywide Wireless Works Fine Here

Its too bad that people don't learn from their own mistakes or the mistakes of others. Likewise, its even more of a shame that they do not learn from the successes of others.

I live in Lawrence, Kansas [wikipedia.org], and just about anywhere I go within the 28.7 square miles of city, I can get a wireless signal from our local wireless ISP, Lawrence Freenet [lawrencefreenet.org]. The service is dramatically cheaper than the local cable company, and speeds are equivalent to DSL. There are routers on lightposts that you can communicate with either directly from your computer (if the signal is strong enough) or you can use a wireless bridge (which they rent and sell) to ensure a good connection. To top it off, you aren't fixed to one location with their service, you can take a laptop with you anywhere in the city, and if you see an access point, you can sign on. How much money did the city spend on this service? Zero. Nothing. This was completely financed by people who believed in it, and that is why it has been successful. With over 1300 current customers after only two years of existence, they are certainly doing things right.

Oh, and I'm posting this from a laptop connected to a Lawrence Freenet access point.

Sad, but not a surprise

When even Google pulls back, it's a bad business model even with advertising-driven models.

Face it, 802.11 is a LAN technology, not a MAN technology. Lipstick on that pig, even with cool mesh network attempts, isn't going to make it better. It was designed for local radial-cellular access by its channelization, and it's not good for covering wide areas. My sentiments go out to Strix and Firetide; both have decent models to make it wider. Cities have to figure out that broadband access is a utility, not an option.

Re:TANSTAAFL - where muni access makes sense

There ain't no such thing as a free lunch.

True, but think of it as "city trash pickup". Of course, a city could just require by law everyone to take their own trash to the landfill like some rural areas, but if it was deemed important enough they could instead have a referendum to raise a tax (sales, property, or income) to pay for weekly trash pickup.

Now some city dwellers may balk at the idea because they can just take the trash to the landfill themselves and save the tax money, but most cities don't work like that so their only recourse is to fight the referendum or city hall legislation. If it passes, the trash pickup truck comes once and week and picks up garbage on the sidewalk regardless of who put it there (could be someone out of town though probaly illegal if they did it regularly) and pick up any amount (some cities have limits of what they pick up but they generally don't go out and report if you left one bag or 5) and they still come by your street even if you didn't put out a bag.

The same thing with Wifi, if a city deems it a reasonable service they can levy a tax and put it to a referendum. If the citizens don't like it they can vote against it, but if it passes then its just like garbage pickup. Not everyone will use it like everyone else (if at all... personally some weeks I'm hard pressed to even put a single trash bag out but I still pay the city tax for trash pickup) and its also going towards out of towners coming into the city and throwing their cups into public trash cans.

So no... Its not a free lunch, but its fine if a city wants a service paid for by tax. You don't see Waste Management complaining that the government holds a monopoly on trash pickup services. Now the service is most likely subcontracted out to the lowest bidder, but in effect it is a monopoly for that city.

The problem with most of these city ones like the Earthlink one in Philadelphia is that they are treating it and expecting income like a regular ISP which intends to make a profit.

It would have been better if Philadelphia had a referendum to use some of the city wage tax to go for free wifi and then explained to Earthlink how much of the taxes they would get rather than expected income from monthly services from individuals who may or may not cancel their account next month.

Re:TANSTAAFL - where muni access makes sense

Except that there is no penalty for everyone else if you don't provide tax-funded WiFi. If you don't collect the garbage, there will be a certain group of individuals who don't take their trash to the landfill, but just throw it out the window (see: 12th century Europe). So far, there doesn't seem to be a penalty to the city for not providing WiFi. That COULD change in the future, but it seems unlikely.

Most taxpayers don't want to see their money going to subsidize a few people who want to use a laptop in the park, even if that's not really the point. And using anything other than a property tax that included business property wouldn't be fair to people who don't live in the area (you mentioned a wage tax - which is almost always a death sentence for local job growth in a struggling economy).

If these municipalities were really serious, they would partner with someone who already has a wide area network in place (like Verizon or Comcast in the case of Philadelphia). For them, it would be incremental revenue, not primary. Pay them enough to get a couple of techs trained to maintain the system, give them some manner of exclusivity and limited liability, and DON'T make them provide end-user tech support. However, the muni could (and should) demand coverage minimums, QOS/uptime requirements, "openness" etc.

Of course, that would require one to admit that networking is hard, expensive, and low margin.