Wireless T1- Innetix Leads Attack On High Costs, Bandwidth Bottleneck
[[ Thanks to DreamWeaver on the WCII thread. ]]
By Mary Eisenhart Photos: Tustin Ellison
In 1982, while still in the Navy, Kenneth Jackson promoted a successful break dancing event in Aloha Stadium in Hawaii and promptly found himself bitten by the entrepreneurial bug. A few years later in the mid-'80s, as an engineering student at San Jose State University, he learned of an emerging opportunity and was quick to act.
"The FCC was having a lottery, issuing cellular licenses for all the major cities across the United States," he recalls. He immediately gathered friends and resources, and "we applied to some of these cellular markets and participated in the lottery." He applied for licenses in 80 cities, and was successful in acquiring an interest in 40 of them. For the next several years, he continued to apply for licenses in a variety of wireless markets--paging, cellphones, etc.--and to ultimately resell them to larger providers.
"Eventually," he says, "I sold my last major license to a company in Chicago. At that point I had the money to either retire or do something else." After a short period of well-deserved R&R, he found himself ready for his next business challenge.
Correctly figuring that the Internet market was about to explode, he produced a local cable TV program called Internet TV, which gave him "the opportunity to interview all the new movers and shakers in the Valley, trying to find out where this whole Internet thing was going.
"One of the common denominators that kept coming up in our conversation was bandwidth, being able to deliver content in a timely manner to people. So that's when I figured, 'Well, why not provide bandwidth wirelessly?'"
The result was San Jose-based Innetix (www.innetix.com), an Internet Access Service Provider that's been offering high-speed wireless Internet connections to its customers for a year and a half, using spread-spectrum technology with 1-watt radios operating chiefly in the no-license 2.4GHz frequency and delivering 1-megabit-per-second speed
in each direction.
While the technology itself is not news (see, for example,
microtimes.com and microtimes.com,
it has been met by a chorus of naysayers who doubt its robustness and scalability (see, for example, www.microtimes.com/debate.html). However, driven by the need for ever-greater bandwidth and frustrated by up to three-month waits for wired connections from the phone company (to say nothing of the monthly charges for those connections), customers found an attractive alternative in Innetix. Not only were they up and running within days (if not hours) of placing their order, Innetix's strategic location across the street from MAE-West and its co-location of servers at AboveNet (www.microtimes.com/165/colocation.html) offered direct backbone connectivity for lightning-fast performance.
Since opening its doors, Innetix has acquired an impressive client list in the South Bay, including software companies, the San Jose Convention and Visitors Bureau, the Fairmont Hotel, the San Jose Black Chamber of Commerce, and Joint Venture Silicon Valley.
Jackson, who has financed Innetix's growth from the proceeds of his previous ventures, says he'd ultimately like to take the company public. Meanwhile, he expresses some frustration that its ability to expand is constrained by the fact that the service provider business model doesn't attract investment from the financial community, even though the same investors pour capital into the equipment vendors from whom the ISPs buy.
"We've got people who pay for the service. We've got a customer base; it's growing; we get inquiries every day; we're still signing people up. So we're beyond the theoretical," he says.
"But before this whole wireless thing can take off," he continues, "the financial community has to step to the plate. Right now the financial community has focused on the equipment side of things. Well, you can have all the equipment in the world, but if you can't set up an office service for somebody and get paid for it, all that equipment you have really means nothing.
"The financial community is going to have to be open to service providers, providing those enabling technologies that are going to allow wireless to spread out."
--------------------------------------------------------------------------------
ISPs being pretty thick on the ground in San Jose, what brought you your first customers?
The mayor of San Jose set up a number of software incubators downtown to spur business development. Initially, Pac Bell was providing T1 connections to a number of them at a reduced rate for a year--and after that year they were forced to start paying the normal rate. A T1 connection was going to cost them about $1500 a month, whereas we're able to provide the same bandwidth for half the cost.
For the incubators, cost was a major consideration.
How would you assess the reliability of the technology compared to wired lines?
I'd say the reliability of the radios is quite good, just as reliable. When we have issues, it's not with the wireless; it's usually with the wired networks in the configuration. We really haven't had problems with the wireless side of it.
You've got the advantage of extreme proximity to the Internet backbone, with MAE-West across the street and your co-location relationship with AboveNet 'round the corner. How does that work out in practical terms?
What we've found in the marketplace is that customers are much more sophisticated in terms of their connection. When a company is signing up with an ISP, a lot of the time they're going to be running mission-critical applications, so they want you to explain a little bit about your connectivity to the Internet.
Because of that, I made the decision early to co-locate our network at AboveNet and at the actual MAE-West, because it allows multiple connections coming in, multiple pipes for big connections coming in from the Internet. It gives us multiple levels of redundancy, and we're one hop away from the Internet, which means that the end-user is going to receive their information that much quicker.
You're delivering services to people a fair distance from your offices, and delivering over distance has always been one of the problems with wireless. How have you solved those problems?
We've had to not listen to the sales hype and the marketing brochures put out by the radio manufacturers. We've had to go out and field-test the equipment and measure the throughput, and then, based upon the need, we provide that specific radio solution.
We've actually had to go out and learn by trial and error what works and what doesn't.
When we find something that works, we leave it in place and never move it.
What's likely to affect the radio's ability to work or not?
The distance. For example, if you try to push a radio beyond its listed distance requirements, you're going to have issues. Even if you put an amplifier on that radio, you could still have issues.
I've heard from friends in the wireless business that part of the difficulty in making this technology work stems from the fact that the radios were often designed for use on corporate campuses and never intended to be deployed in Metropolitan Area Networks. Is this the case here?
Initially they were intended for corporate campuses, but then they started increasing the power and adding amps to it. As long as you stay within the distance limitations, you're okay. Once you start going beyond that, you start having problems.
The challenge has been, how do you configure in such a way that you're not going to violate those distance issues, while at the same time having a customer base within the general vicinity so you can operate a business and make money.
So, for example, the Breezecom (www.breezecom.com) radios, which go about three miles, allow us to hit pretty much everything within the downtown area and things close to downtown. When we go beyond three miles, we look at a different radio solution.
One of our prospective clients, Mission College in Santa Clara, requires a 3-meg-per-second connection. We had to shoot that from AboveNet to Mission College through the San Jose Airport. Even though we had radios that said they could do four megs, by the time we put it up, and measured it on the Mission College end, we were only receiving about 1.2.
What is the problem the airport poses?
The only thing we could think of was interference, but most of the frequency the airport uses is 8-900MHz. We couldn't understand whether it was the distance, the frequency, or what, causing the problem.
What we did was go to a different frequency, and got a higher-bandwidth radio. The radio we have now will deliver six megs, so we'll meet their criteria.
What's the power of these radios?
One watt.
And how long a distance can you go between repeaters?
It depends upon the radio. Some of them will go as far as 15 miles, but for us, 15 miles is good enough, because we primarily serve customers within a 10-15 mile radius.
That's where the majority of the business is, here in San Jose.
We have a number of customers in the Fremont and Milpitas area--we started with about 10 customers in Fremont. We needed a backbone connection going up to Mt. Allison to be able to serve those customers down in Fremont. In that situation we'd put a higher-bandwidth radio, like a 6-meg radio, as the backbone. It would feed the mountaintop, then from the mountaintop, we'd slice that up and provide bandwidth at some measured rate to the end-user.
Mt. Allison allows us to go into Fremont, Milpitas, the whole Newark area, all the way up in there. Our radios on Mt. Umunhum allow us to service Campbell, Saratoga, Cupertino, Sunnyvale, and Palo Alto.
You're starting to offer service in Oakland. How is that being done?
We've run a DS3, or a big pipe, going up to Oakland. We've secured one of the tall buildings in downtown Oakland, and now we're in the process of actually bringing the bandwidth up, wiring it up to the roof so we can beam from the rooftop. Already we've gotten requests from Emeryville and a lot of places in Oakland where they can't receive wired T1 service.
We would like to offer entire-Bay-Area coverage. With our existing infrastructure, anybody who's an ISP end-user all the way up to a T1 user is a potential customer for us, even with other technologies like XDSL and ADSL on the horizon. Just the other day I got an email from somebody about ADSL, saying that it only operates within three miles outside of their Central Office, so they still need a solution.
If you're in the wrong place, you just can't get it.
Exactly. Even though there's cable modem and ADSL, and that's fine, there are still a lot of holes in the infrastructure. So there's going to be room for wireless.
The way it's evolving now is that a lot of ISPs have contacted us about partnering opportunities. What we're trying to look at now is how we can bring other ISPs under this umbrella--how do we franchise it? How do we allow other ISPs to add this as part of their value-added package, and how to package this all together? Right now I'm talking to a number of ISPs about that very concept.
People have been working with this technology for some time, but why is it suddenly ready for prime time?
The reason I believe it's ready for prime time is that, especially for us in the downtown area, it's a solution that's quick. They can call us up, we put an antenna in their window, and they're up within a day. They're paying half the cost, they don't have to get in that Pac Bell queue and wait months and go through that whole process.
If you're in a 10-15 person office and everybody's accessing the Internet, you'll truly be impressed by the speed versus your regular dialup connection. These are people who are used to surfing the Internet, tired of slow speeds--they want to get on, get the information, and get off, and they're looking for much faster access.
So I think the big issues are the time savings of getting them up, the cost savings--and the fact that it works!
Copyright c 1998 by Mary Eisenhart and MicroTimes. All rights reserved.
Sidebars:
Wireless: Coming To An ISP Near You?
Wired Vs. Wireless--Comparative Cost Analysis |
