By Jeffrey Krauss, Satellite tracker and President of Telecommunications and Technology Policy

Later this year, Iridium's satellites will be "de-orbited," and they will burn up as they fall. This column explains what went wrong, and the implications.

Mobile satellite service

Iridium was designed to offer "mobile satellite service," or MSS service to mobile terminals. There are several other MSS systems operating, including Inmarsat and American Mobile Satellite Corp. (AMSC). They both differ from Iridium in that their satellites are geostationary—they orbit at fixed slots above the equator, 22,000 miles up. Iridium satellites go around and around at an altitude of 485 miles. That difference in altitude means that Inmarsat and AMSC terminals must use parabolic dishes pointed at their respective orbital slots, while Iridium supports much smaller handheld terminals that look like the cellphones of a few years ago.

Iridium was designed in part to satisfy the whims of wealthy tourists and traveling businessmen. While the Iridium system apparently worked OK, it took so long to come to market that terrestrial cellphone service was deployed and available nearly everywhere wealthy tourists might want to travel. Motorola announced the Iridium program in 1990, but service didn't start until 1998.

And Iridium was designed for voice service; it's not really suitable for Internet access. Although it used to be that a digitized voice circuit required a data rate of 64 kbps, digital voice coding has improved so much that only 8 kbps or less is needed to carry a voice conversation. While this is efficient use of bandwidth, you can't send very much Internet data through an 8 kbps pipe.

Iridium needed this level of bandwidth efficiency because the FCC gave Iridium so little spectrum. People forget (or maybe never knew) that Iridium was licensed to operate only in the 1621.35–1626.5 GHz band, or slightly more than 5 MHz of spectrum. This happened because there were originally four MSS applicants, Iridium, Globalstar, Ellipso and Constellation, and the FCC had to figure out a sharing arrangement for the 1610–1626.5 and 2483.5–2500 MHz bands that they wanted to use, a total of 33 MHz. Globalstar, Ellipso and Constellation were able to share with one another because they each employed a CDMA design, but Iridium was based on a TDMA design and was unable to share with anyone else. So the FCC divided up the pie, and Iridium got somewhat less than its fair share.

Now, Globalstar is operational, but Ellipso and Constellation never found funding, so Globalstar doesn't have to share with anyone. Iridium would have had a greater chance of success if it had access to more spectrum, because it would have had a chance of converting from a voice service to an Internet access service.

The future

It's too soon to say whether Globalstar will be successful, although its costs are reputed to be lower than Iridium, and its marketing plan is not pointed at wealthy tourists. Meanwhile, there's another round of MSS systems in the works. ICO, a spinoff from Inmarsat, and a bunch of other companies have announced plans to operate in the 1990–2025 and 2165–2200 MHz bands (a total of 70 MHz). These satellites, like Iridium, will orbit at low altitudes and work with handheld terminals.

ICO is the only one that has made any progress, and it hasn't been very impressive. ICO contracted for the construction of satellites and then couldn't raise the cash and had to file for bankruptcy. ICO was then acquired at fire sale prices by Craig McCaw, a neighbor and sometimes partner of Bill Gates, and an expert in wireless communications. McCaw and Gates are partners in Teledesic, a planned low earth orbit satellite system operating with fixed earth terminals (at customer locations) and much more spectrum (a total of 1000 MHz), but at much higher Ka-band frequencies (where rain attenuation is a serious issue). McCaw actually kicked in some money to keep Iridium alive a few weeks longer, considering whether to fold it into ICO, but abandoned that plan when he saw how little bandwidth Iridium had.

McCaw clearly plans to use ICO as a mobile adjunct to Teledesic for wireless Internet access. The current generation of cellphones, while great for voice, can't support the higher data rates that are needed. Next-generation high data rate cellphones are on the way—Qualcomm has a design called HDR, and Motorola calls its CDMA 2000 1X Plus—but a standards battle may slow down deployment.

Meanwhile, the other companies seeking to operate on the same frequencies as ICO—including Boeing as well as Iridium, Constellation, Ellipso and a bunch of others—haven't been able to arrange any funding.

So we will be left with two MSS operators, Globalstar and ICO, operating on different frequency bands. And maybe that's plenty, considering the deployment of cellphones. Having a great system design, and having a license from the FCC, is not always enough. You need customers, too. And you need to get your product to market fast, before your competitors grab all the customers. Otherwise, like Iridium, you might crash and burn.