Carolyn et all, an article from Oristano the head of SpeedChoice, division of People's Choice.
businesstech.com
We are a high-speed-access service. The fact that we're wireless is great, but the fundamental market we're in is the high-speed access service. And in that market, at the moment, the major players would be, of course, the cable companies, with cable modems; the telephone companies and their resellers, with ADSL; and then us, in the markets we're in. And, to a much lesser extent, the satellite-delivered services, like Direct PC. We really haven't seen a lot of market share out of them.
And so when you think about it, for instance, Phoenix -- the market we're launched in now -- if you are a homeowner in Phoenix, you have three different people you can call who are selling a high-speed-access product. The issue you're going to run into in Phoenix is, the cable company covers about 30% of the market. They may say it's more, and they may have rebuilt some more. But, I think, last I heard, it was about 30%. And ADSL, although they'll take your call from anywhere, we don't think the coverage is much more than 25%. Whereas the Speed Choice coverage in Phoenix is 90%-plus. So we see ourselves as competing in the market for both commercial and residential high-speed access.
BT: Let's, for example, take Phoenix and look at the line-of-sight transmitter issues that might be affecting the customer base, and what you can do after the initial rollout to capture more customers that aren't served right now.
MO: Okay. Well, in Phoenix, to give you a sense of the way the market is currently configured, we have two cell sites currently active; we have an engineer who calls them "super cells," for lack of a better phrase. What we mean by that is, they're really big. The South Mountain cell site has a 35-mile-plus radius, so it's not small, mobile-type cells. There is another cell on a mountain sort of in the middle of the market which casts a shadow -- and we project behind that to about 70,000 homes who otherwise would be in shadow from the first site.
With those two cells -- which are both currently active -- with those two cells we have about 90% line of sight. So the question would be: What are the economics for going after the other 10%? And I think, right now, we feel we need to have more time doing installations in the market before we feel we know exactly where that other 10% is, to go put in ever smaller and smaller cells. But, ultimately, I think you'll see us go add more capacity in that way.
So right now we can hit 90% of the homes in Phoenix. And, generally, the way they are hit is with installations where the antenna -- which is about one square foot in size ... it's about 10 to 12 inches on a side -- that antenna is either on the side of the home or on a very short, three-foot pole. It's generally not higher than that.
MO: I've kind of said that's a bit of a high-class problem for us, because we have such a large amount of spectrum to dedicate to the service. So if we get to the point where we have so many customers that there's a drop-off, that would be a very large number of customers. Let me run you through a couple of the basic sort of numbers, the way this works.
The downstream spectrum is, say, 30 channels of downstream spectrum, and that's 6 megahertz each. That's 180 megahertz of downstream spectrum. We will use no more than half of that for video -- you don't need to with digital compression. So let's say we use half of it for Internet. So we're using 90 megahertz of spectrum for Internet. You can then go and take our cell sites, and say: Okay -- well, we've got two cell sites, so it's not 90 megahertz and that's it. It is, in fact, 180 megahertz, 'Cause we've already doubled it.
And then you can look at sectorizing it. And we haven't had to get to that point, but sectorization of the downstream to 30-degree sectors is not terribly difficult. And a 30-degree sector, if you're deploying it against 180 degrees -- which is sort of a semicircle pattern -- would give you a multiplication factor of at least three times reuse. So I would have 180 times three -- I would have 540 megahertz of actual spectrum.
Our transmission efficiency is 5 bits per hertz, -- the same thing that the standard cable modem uses. So multiply the 540 times 5, and you've got 2.7 gigabits of capacity, in a kind of basic two-cell-system design.
Now you say, if you want people to be able to download -- at a megabit, at the worst time -- we've got 2,700 megabits of general capacity. You would have 2,700 absolutely simultaneous downloads? Instantaneously, from that design. I've got to tell you -- when you start looking at the kind of ethernet statistics that come out of this technology, you find that the percentage of instantaneous downloads is very, very low. It certainly is on the order of no more than 10%. So that would mean 10% of all the people who are on line might be downloading at the same time, at the worst.
So if you divide that by 10%, what you're saying is you can have 27,000 people using the service simultaneously, before you degrade below a megabit per second? But wait a second -- that's people using the service simultaneously. How many people are actually sitting in front of their computer using the service at any given moment? Well, the average ISP has a modem-to-customer ratio of about 10%. Let's say we get much more intensive use by our customers, and they're sitting down a lot more, because our service is better.
Say it's 20%. So you take the 27,000 people, you divide it by 20% -- you end up with a basic capacity of service for 135,000 customers, in the spectrum, before you have to think about adding a third cell site. It ends up that we have so much spectral capacity, before we have to deal with the question of even adding more cell sites. It's a problem that I really relish dealing with, because it'll mean we have an awful lot of customers.
Its a long article but worth the reading.
Hiram |
