Fixed Wireless Business

Mesh At Work

Mesh network architecture has only recently begun to gain mindshare among equipment vendors and service providers, but one pioneering ISP in northern California is already up and running—and earning revenue—using mesh products.

by Gerry Blackwell [March 12, 2002]

Mesh architecture solves line-of-sight problems by using customer sites as relay points, allowing networks to route around obstacles. Because of their non-line-of-sight capabilities, mesh networks are being touted as a way for service providers to economically attack residential markets.

That's exactly what Vista Broadband Networks Inc., a start-up ISP in Petaluma, Calif. north of San Francisco, is doing. Vista is using the first mesh product to market, Nokia's 2.4GHz RoofTop Wireless Routing solution.

"We wanted to get to the broadband residential market without paying the telco," explains Vista Chairman, Chief Eexcutive Officer and Co-founder Scott Mindemann. "We tried many architectures, but eventually we came to the conclusion that the way to do it was with mesh."

"Mesh allows you to use standard technology radios to route around obstacles and give ubiquitous coverage. There is nothing more efficient to get to residential neighborhoods."

From MMDS to mesh
Although Vista has only been in operation officially since October, the core management team, which started in the Multi-channel Multi-point Distribution System (MMDS) business, has been together for ten years, Mindemann says. And it's been testing the Nokia technology since May 2000.

At the time the Vista brains trust decided mesh was the way to go, Nokia RoofTop was the only available mesh product. A few other vendors now have products on the market and many more are coming, Mindemann says, but the Nokia gear is still the most economical.

Today Vista has the largest existing commercial network using the Nokia products—though it's small yet: just 100 residential customers scattered along a corridor that runs through the San Francisco bedroom communities of Novato, Rohnert Park and Santa Rosa.

This is still impressive progress given that the company only launched its first serious marketing campaign in the second week in January 2002, Mindemann says.

Customers pay $45 and up a month plus a $200 account initiation/install fee. The entry-level package delivers data rates up to a DSL-equivalent 384 Kbps. It's symmetrical and always on. Customers actually experience throughput of somewhere between 300 and 600 Kbps, Mindemann says. Vista can offer higher data rate packages as well.

Mesh math
Mesh makes sense for residential markets for a bunch of reasons, he says.

First of all, of course, it solves the line of sight problem. With only five or six customers installed in a neighborhood, Vista can be sure of at least 90 percent coverage, which is a far cry from point-to-multipoint technology.

This has important implications for the cost of customer acquisition. When service providers with point-to-multipoint technology market to a neighborhood, they can only be sure of being able to service 15 percent of the customers who respond, Mindemann says.

This means much of the marketing materials and pre-sales customer interactions are actually wasted. And propagation anomalies make it difficult or impossible to predict which houses in a neighborhood will be able to get service. So the company also has to send a truck out to find out.

Of course, few if any service providers with conventional point-to-multipoint technologies are attempting to crack residential markets for exactly these reasons.

"From the acquisition costs perspective, mesh is hands down the best," Mindemann says.

Getting the ball rolling in a neighborhood is slightly "awkward," he admits. You need a half-dozen customers at least to guarantee more or less ubiquitous coverage. So when the very first customers call in, Vista can't commit until it gets a few more.

It's usually not a problem. In one recent case, the company identified a target neighborhood of about 220 homes. It launched a very local marketing campaign with hopes of attracting 40 customers to make it economically viable.

Because the company had selected the area carefully for demographic profile, it was able to sign up 28 customers in a week. At that point it had no infrastructure in place, but felt comfortable promising service within a month. In fact, it took six days to hook up the first 40 customers.

"One month later," Mindemann says, "we were in a revenue generating position."

Mesh truck rolls
The other huge benefit of mesh is that while it does require a technician to install the system at each home, truck rolls are much more efficient. With point-to-multipoint equipment, a technician has to aim the antenna and may have to install it on a mast.

With the Nokia RoofTop equipment, Vista "just bolts the antenna on the side of the chimney and the network comes up right away," Mindemann says. There is very little skill needed and it's very quick."

Besides, "a truck roll isn't a bad thing sometimes if it's done properly," he adds.

Vista takes the opportunity of selling customers home networking gear and installation services. It's finding a steady demand for Wi-Fi wireless, telephone line and even conventional wired Ethernet networks.

This market will doubtless tail off as customers become more savvy and realize they can buy gear at retail and install it themselves, Mindemann says, but in the meantime, "the truck roll becomes a revenue generator."

Mesh markets
Vista is mainly targeting telecommuters, home offices and technically advanced families. Its strategy is to carefully select neighborhoods in urban areas with populations of between about 30,000 and 150,000 where neither DSL nor cable is available.

It looks at communities near big cities with a high proportion of professionals and vibrant business economy.

Once selected, infrastructure costs to get a neighborhood up and running are reasonably low. Nokia RoofTop customer premises equipment costs about $800. The Airhead, the neighborhood hub device, sells for about $1,200.

On the face of it, that's more expensive than point-to-multipoint. "But the exciting thing about mesh is that in effect you're funding infrastructure with each customer you add," Mindemann says.

Then there is the cost of interconnection—either a T-1, multiple T-1s or, in Vista's case, often a wireless backbone link.

Vista is also offering business Internet access services—T-1 to 10 Mbps—using conventional point-to-point and point-to-multipoint wireless technology. It has about 100 business customers.

One key strategy is to find residential customers—telecommuters in particular—through business customer contacts. And then use point-to-point infrastructure established for business customers to backhaul residential traffic.

More to mesh
Even though it's using the Nokia equipment exclusively on the residential side and is convinced that mesh is the only way to go in that sector, Vista is completely technology agnostic, Mindemann insists.

For business customers, it's currently using Adaptive Broadband gear from Axxcelera Broadband Wireless Inc. of Santa Barbara, Calif.

But Vista is also testing products from "just about everyone else making wireless point-to-multipoint equipment." It's particularly impressed so far with products from Milpitas, California-based Aperto Networks.

One senses the real game for Vista is the residential market. While it is committed to mesh and Nokia now, it will try other mesh products as they appears, and even, Mindemann says, other non-mesh NLOS products.

But for now, he says, mesh is hard to beat.

—End

Online resources:
  ѷ Fixed Wireless How-To Handbooks