Carrier WiFi

Rainbow Unveiled

As we predicted (see Rainbow on the Horizon?), Intel Corp., IBM Global Services, AT&T Corp. (NYSE: T), and others have today announced the formation of a company that will build out a huge network of 802.11 hotspots in the U.S. and sell wireless Internet access to carriers, ISPs, and other service providers. The company is called Cometa Networks (see Cometa Plans Nationwide WLAN).

Privately held Cometa is the outcome of the initiative called "Project Rainbow," which has been causing quite a buzz in the U.S. WLAN industry since details about the project first leaked back in July (see WLAN USA? and Rainbow to Link WiFi & WAN). "Cometa is Italian for Comet," the company's CEO Larry Brilliant told Unstrung. "So the rainbow has now become a comet." Brilliant!!

Whatever you care to call it, Cometa's WiFi buildout plans are an order of magnitude more ambitious than anything that has been attempted before. "We intend to have a Cometa hotspot five minutes away from anyone in a major metropolitan area in the U.S.," says Brilliant.

He estimates this will require between 25,000 and 50,000 hotspots, maybe more. The largest hotspot rollout so far announced in the U.S. is T-Mobile USA's plans to have 2,000 access points up and running by the end of this year.

Brilliant says that Cometa will put the access points in restaurants, cafes, electronics retailers and other stores, and hotels. "Airports are certainly not off our screen [either]," he says.

"We will start in the top ten MSAs [metropolitan statistical areas] in the U.S. and then work our way through the top MSAs in the country," says Brilliant. Installation will start immediately, he adds.

IBM Global Services will be installing hotspots for Cometa. The company is not revealing its preferred equipment providers. Brilliant would only say they will be "best of breed" and support both the "a" and "b" variants of the 802.11 standard.

The firm plans to offer wireless LAN access through its network to pretty much any service provider that wants it. "Cometa owns the network and access is resold by carriers, ISPs, and other providers."

AT&T will be the first major carrier to offer services using Cometa networks, and Brilliant says the operator is planning to use the network to offer wireless services to its installed base of corporate VPN users. There are, he says, 10 million such users in the U.S., and AT&T has about 40 percent of that market. Corporate VPN users will be the initial target of the wholesale WLAN service, Brilliant says.

What about the backhaul required to connect the hotspots to a broadband Internet connection? "AT&T will be providing adequate backhaul in every location... [exactly how much] will depend on the location." AT&T says that the service will use the AT&T IP backbone network. However, the operator will give no estimates of the cost of providing bandwidth for 25,000-plus hot spots.

So what's in this for service providers? Buying in access from an existing hotspot network without having to build out their own could bring them "significant margins," according to Ted Schell, chairman of Cometa and a general partner at VC firm Apax Partners, which is backing the venture along with 3i Group plc.

Despite the talk of money and margins, no one at Cometa is happy to talk about how much cash is being pumped into the venture. "Let's just say it's more than Ted's salary, and you know these VC guys!" Brilliant quips brilliantly.

However, he did offer a brief glimpse into money matters when Unstrung asked if it was actually a bit late to start planning a WLAN network, when there are already so many other small-scale providers out there. "No one is putting up tens of millions of dollars before the demand curve here," Brilliant rejoined. "Are we too late? No, I think we're right on time."

The management is also not keen to reveal the makeup of the company, in terms of what portion of the venture is owned by each of the partners. "I think one of the disappointments of this conversation is that we can't reveal these details," Schell told reporters after being pressed on the issue on the conference call that officially launched the company this afternoon.

The other big name in this partnership, Intel, is providing technology consultation for the venture and tying it to the launch of its Banias WLAN-compatible processor in the first half of next year. The chipmaker will optimize the radios on the chipset for the Cometa network, according to Mark Christensen, VP and director of Intel Capital's communications sectors.

To be honest, we're not sure how you fine-tune a component that is supposed to entirely conform to the Institute of Electrical and Electronics Engineers Inc. (IEEE) 802.11 standard. Perhaps Intel has a home-grown data transfer speed boost mode up its sleeve for Banias? Stranger things have happened...

Anyway, readers will no doubt be glad to know that you don't have to use a Banias chipset to access Cometa's network -- they'd just prefer it if you did. The hotspots, says Cometa, will be accessible to anyone that has a standard 802.11 card.

— Dan Jones, Senior Editor, Unstrung
Page 1 / 2   >   >>
lrmobile_castro 12/4/2012 | 9:14:04 PM
re: Rainbow Unveiled Wi-Fi is great and it can do a lot of things well. Some of those wi-Fi hot spots are also very useful. But i am the only one that thinks that the Wi-Fi bubble is clearly obstructing common sense?

It has a place, even in the public market, but it being extended beyond economics. Wi-Fi the new 3G.
spc_myles_telos 12/4/2012 | 9:13:45 PM
re: Rainbow Unveiled Wi-Fi is a proven access technology. Extremely easy to put up and not exactly very expensive either. As a disruptive technology, it's become gradually accepted and you will definitely see more of it.

The media will continue to make it a big deal. I wouldn't go as far as to say that Wi-Fi is the new 3G. The amount of Wi-Fi $$$ dollars circulated from vendors to SP's, extremely lower!

untethered 12/4/2012 | 9:13:35 PM
re: Rainbow Unveiled Those of us "in the know" don't see as many $$s being spent and don't see it as a big deal - isn't this the very nature of disruptive technologies though?

I think that while the front end(radio/access) costs of WiFi are lower, the back end (OSS/BSS, support, backhaul) costs are just as obscenely high as any other access technology, once we get over the idea that the only people using this will be technology-savvy folk utilizing "thick clients" (i.e. laptops and PDAs).

AT&T is probably chasing the "Digital One Rate" model all over again (AT&Ts "seamless" network was a very big factor in jumpstarting the US wireless market in the late 90s).

joset01 12/4/2012 | 9:13:31 PM
re: Rainbow Unveiled unthethered wrote: "the back end (OSS/BSS, support, backhaul) costs are just as obscenely high as any other access tech"

Yeah, its very hard to get anyone to say anything useful about backhaul costs for this project. AT&T say they will connect the hot spots to their IP network, that's about all.

Does anyone have more info/estimates on backhaul costs for this project? If so, post here or email me on [email protected] Thanks.

DJ Unstrung
lrmobile_castro 12/4/2012 | 9:13:29 PM
re: Rainbow Unveiled To me this is more proof that the carriers really have no idea what will work. I think Wi-Fi is the next 3G and a huge bubble. Here is why.

When 3G was initially promoted, carriers, vendors, and analysts suggested nearly limitless sources of revenues. yet now that carriers have seen the reality of 3G's business model, they are backing out and/or slowing down.

Wi-Fi is in the same place. It is the pancea for all. Yet few have made money from it and i doubt that anyone really knows what the business model is like when it scales.

Next step, carriers and others will learn the wi-fi model and start closing hot spot points. Its nice to see carriers exploring different options such as wi-fi, imode, 3G but they have yet to hit the one that is really going to help them. Additional exploration is required. Frankly that answer is probably a combination of approaches.
spc_myles_telos 12/4/2012 | 9:13:29 PM
re: Rainbow Unveiled You can use groomed Cable/DSL for this kind of thing. I mean, depends on what level of committment/service you expect. Mobilstar's network was T1 groomed into an ATM DS-3 network I think... standard service provider WAN stuff...

I know here in BC, you can get a service called E-ADSL... where they groom everything (ADSL as access) into an ATM network, do whatever you like.

If you use gear from a "secure AP vendor" like if they go with something from Colubris, I believe they could even backhaul via the net directly. T1 access to the net should be cheap.

Use VPNed backend to authenticate users.

T1's to the net are what... <1000/month by now i would imagine?

I wouldn't be surprised if Cometa also partners with companies, do a half-MVNO model.

Lots of options, I'm sure they have a few actuaries in the back, figuring it all out... and maybe even bandwidth companies lurking to join the party...


joset01 12/4/2012 | 9:13:27 PM
re: Rainbow Unveiled Well, if you remember they were originally talking about roaming between between different network types. IBM has an app that will maintain the connection (they say!) but apparently this isn't now part of the initial rollout.

The only way that WiFi really makes sense to me is as a free or cheap internet access method (usually offered by individual stores etc) OR as part of a carrier's offering (that can be secured and paid for on one bill and maintain a connection accros diff network types).

I think there are a bunch of technology problems around WLAN for carriers e.g - very few WLAN/WWAN chips or chipsets available, no voice over IP (VOIP) (see http://www.unstrung.com/docume.... However, I guess that there will be a bubble (see http://www.unstrung.com/docume... - at least until people realize that WLAN isn't a panacea.

DJ Unstrung
reg64 12/4/2012 | 9:13:23 PM
re: Rainbow Unveiled I tend to agree with the posts that lean towards the difficulties of the carriers realizing any substantial recurring revenue from Wi-Fi. However, it would seem that the equipment vendors would be in decent shape longterm even with the prospect of declining hardware prices.

The infrastructure segment of Wi-Fi should continue to see large growth for years to come just by virtue that the technology represents a huge cost avoidance over a hard-wired solution.
Why continue to pay huge bills for adding and moving data wiring whenever there is need to expand the premises or add new employees? With the Wi-Fi security issues now being adequately addressed, this could be the segment that experiences the more robust growth. Compared to what is encountered with trying to make money off of services, the infrastructure market will be a slam dunk.

I believe it was Verizon that recently announced a plan to build out wireless infrastructure for their small and medium business customers who are using a wired solution. There will be no recurring revenue, but they will make some money with any opportunity to add other services in the future. They seem to have the best initial Wi-Fi strategy.

You would think that the Corning and Lucent (fiber suppliers)are in big trouble.
untethered 12/4/2012 | 9:13:18 PM
re: Rainbow Unveiled Does anybody else think it is omninous that not a single handset (laptop, PDA, phone) vendor is a part of this venture?

Handset subsidy has been a big part of the cellular business model in the US (that and the lack of number portability) and hence a big expense. Without a partner, Cometa has to eat the difference.

On the other hand, look at this from the eyes of the investors. To each one of the 5, this is a relatively minor investment because the spectrum is free and the market has already been seeded with access equipment.

joset01 12/4/2012 | 9:13:17 PM
re: Rainbow Unveiled Depends. This seems to me a laptop-centric venture at the moment. Hence Intel's involvement -- it wants to sell its new chipset.

Now, if there are good WLAN/WWAN chips for phones coming I guess we'd see them in some high-end smartphone hough I dread to think what the battery life would be). But I would think that its probably at least a year too early for this to be really interesting for handset manufacturers.

DJ Unstrung
Page 1 / 2   >   >>
Sign In