The Ka-band (18.8-19.3 GHz downstream/28.6-29.1 GHz upstream): Without a doubt, it’s the next great frontier for VSATs (very small aperture terminals). Offering substantially more bandwidth than either C- or Ku-band, Ka-band seems a natural for large Internet-style data transmissions: the kind of communications that are being increasingly demanded by businesses and consumers alike.

However, this new spectrum comes with a price tag attached, and that’s rain fade. Because it operates at higher frequencies than C- or Ku-band, Ka-band broadcasts are more vulnerable to being washed out by heavy storms in their path. That’s why vendors are only talking about 99.5 percent guaranteed up-time for the Ka-band, as opposed to the more than 99.9 percent availability associated with the C- and Ku-bands.

Is this a problem? Well, that’s a hot topic of debate in the satellite industry. But in order to understand both sides of the issue, we first need to know more about Ka-band as a satellite transmission medium–and the plans vendors have for it–before asking if Ka-band VSATs are really ready for prime time.

In The Beginning

Before 1993, most people didn’t think the Ka-band could be used for satellite transmissions. However, NASA’s Advanced Communications Technology Satellite (ACTS), which was launched in September that year, changed their minds. Parked in a geostationary orbit at 100 degrees W, ACTS was the focus of many successful Ka-band transmission experiments. NASA used it to test all kinds of platforms for Ka-band transmissions: not just earth stations, but also aircraft, land vehicles and ocean-faring ships. In addition, the variety of applications tested was stunning. Beyond basic communications, uses such as telemedicine, distance learning, and remote monitoring of electrical utility networks were also put through the wringer on ACTS.

To say the least, ACTS was a huge success. During its 81-month run–more than three times longer than planned–more than 100 experiments were conducted with this satellite. The results were undeniable. NASA not only proved that Ka-band could be a satellite transmission medium, but that satellites could send and receive data at 45 Mbps downstream/1.54 Mbps upstream using only 26-inch earth station antennas.

Of course, this ability came with a few strings attached. First was the rain fade problem. For Ka-band uplink earth stations to cut through heavy rain, they have to be capable of radically boosting their signal output at a moment’s notice. According to John Evans, chief technology officer at Comsat Labs, this means that a VSAT earth station usually running at 0.1 watts must jump up to 1.0 watts or more to make itself heard by a Ka-band satellite.

The next consideration is coverage. Thanks to the signal losses caused by rain fade, Ka-band satellites aren’t able to blanket an entire continent the way a C-band satellite can. That’s why the Ka-band satellites currently under development use a number of focused spotbeams to reach their coverage areas. This allows them to cut through the rain, but only at the cost of more EIRP (effective isotropic radiated power).

The upside of this approach is the spotbeams can be aimed wherever they’re needed. The downside is that uplinking information from one spotbeam and then downlinking it via another requires signal switching. This switching has to be handled within the satellite itself, which adds extra complexity to this system. That’s not good news, since extra complexity typically spells lower reliability and higher costs in any engineered system. (Another solution is to forego the onboard switching, and have the satellite only speak to stations dedicated to specific beams, with no crossover between transmit channels).

Satellite equipment manufacturers and service providers, however, are not daunted by the challenges of Ka-band. Far from it. In fact, many have already announced ambitious plans for this spectrum.

The Next Generation

For instance, Hughes Network Systems is plowing ahead with its Spaceway high-speed bandwidth-on-demand satellite system. It’s to be built on a series of HS 702 geostationary satellites with “multiple spotbeams and onboard processors to provide very high-speed mesh connectivity between small terminals,” says Michael Cook, Spaceway’s vice president and general manager. This means Spaceway VSATs will be able to communicate directly between each other, rather than having to be routed through hubs like traditional VSATs do. The result will be faster communications between sites, more flexibility in configuring networks, and fewer pieces of equipment in the transmission chain.

More importantly, there’s the speed of Spaceway: its VSAT earth stations can uplink at data rates up to 6 Mbps–outstripping conventional VSATs by a wide margin–and downlink from the satellite at a mind-blowing 400 Mbps.

With performance like this, it’s not surprising that Cook describes Spaceway as “the next generation of VSAT systems.” After all, it’s been designed “to address the whole range of market opportunities and market segments,” he says. “We’re looking at everything from major corporate enterprises right down to the consumer.”

The launch date for Spaceway’s first trio of Ka-band satellites is set for 2002, to cover North America. Later on, six more orbital slots will be filled around the earth to provide global Spaceway service.

Lockheed Martin Global Telecommunications (LMGT)–with the help of the Telecom Italia Group and TRW–also has plans for global high-speed services via Ka-band. Called Astrolink, the service will use LMGT geostationary A2100 satellites to deliver up to “20 Mbps in the uplink, and 220 Mbps in the downlink,” says Celso Azevedo, Astrolink’s president and CEO. Astrolink’s first satellite launch is set for 2002. Covering 92 percent of the globe, it will start with a constellation of four satellites, with the option of expanding to nine located in five orbital slots.

Like Spaceway, Astrolink’s A2100s will use onboard processors to provide full mesh connectivity between VSAT terminals. However, unlike Spaceway, Astrolink isn’t trying to be all things to all people. Instead, its target market will be “multinational corporations and government, and small to medium enterprises,” says Azevedo.

As for Gilat? Well, this VSAT powerhouse is definitely interested in Ka-band for VSATs, says Dianne VanBeber, vice president of investor relations for Gilat Satellite Networks. In fact, “we are presently working on Ka-band VSAT hardware, and hope to have something available sometime this year,” she says. Asked what kind of services Gilat intends to offer, VanBeber replies, “We expect to use Ka-band not only for our traditional VSAT customers, but also for consumers wanting broadband Internet access as well.”

Without a doubt, the most ambitious of all Ka-band applications has to be Teledesic, the low earth orbit (LEO) brainchild of cell phone mogul Craig McCaw and Microsoft founder Bill Gates. Founded 10 years ago, Teledesic has forged ahead with the dream of creating a global, broadband “Internet-in-the-Sky” using 288 Ka-band LEO satellites. As planned, its network access will come in two flavors: a standard configuration of 2 Mbps upstream/64 Mbps downstream, and a broadband version offering 64 Mbps both ways.

The real challenge for Teledesic will be managing the inter-satellite communications needed to make this global system work. That’s because Teledesic’s LEO architecture requires a constant hand-off of signals from satellite to satellite. The year 2004 is the date to mark on your calendar: that’s when Teledesic’s service is scheduled to begin.

This said, many in the satellite industry wonder if Teledesic will ever get off the ground. One of those people is Roger Rusch, president of Telastra Inc. and a long-time industry observer. Commenting on Teledesic majority owner Craig McCaw’s rescue of bankrupt satellite telephony firm ICO Global Communications, which has been relaunched as the New ICO‚ Rusch says he expects McCaw and his team to focus on that enterprise first and foremost. As for Teledesic? “Nobody’s declared that it’s dead,” says Rusch, “but I think the reality is it’s not likely to proceed.”

Meanwhile, in Canada, equipment manufacturer Norsat International has signed a deal to supply Korea’s LG Information and Communications with approximately 2,000 two-way Ka-band outdoor units (ODUs). Billed by Norsat as “the world’s first commercial order for two-way Ka-band end-user terminals,” the Norsat ODUs will be used for distance learning via Koreasat 3, which has three Ka-band transponders onboard.

“We’ve been involved in Ka-band for years,” says Gordon Deans, Norsat’s vice president of broadband networks, including Ka-band projects with the European Space Agency, the Canadian Space Agency, and ACTS. This includes Norsat’s development of the Pico, a small Ka-band terminal that can be packed up in a briefcase.

On another front, marine satellite manufacturer Sea-Tel has been working with the U.S. Navy and NASA, testing Ka-band out on the waves. In a recent ACTS experiment on Lake Michigan, the Naval Research Laboratory and NASA’s Glenn Research Center achieved a 45 Mbps satellite link to the 45-foot yacht Entropy.

The link, which was maintained by a Sea-Tel tracking pedestal, was a breakthrough by naval standards. That’s because 45 Mbps is at least 20 times faster than the current shipboard standard for satellite transmissions.

Moreover, these results were noteworthy by shipboard satellite standards; that’s because Ka-band demands highly accurate antenna pointing, says Sea-Tel president Pat Matthews.

“A C-band antenna doesn’t experience a 3 dB loss–which is serious–until it’s 0.75 degrees off the satellite,” Matthews explains. “A Ku-band antenna will do the same at 0.65 degrees off. However, a Ka-band antenna has to be pointed within 0.4 degrees of the satellite, or it too will experience a 3 dB loss. As margins go, that’s pretty tight.”

Concerns Remain

This, then, is the current state of Ka-band satellite developments. As Comsys Senior Consultant Simon Bull puts it, what’s happening now is “a lot like a swan swimming. There’s an awful lot of activity happening under the water, but you can’t see it.”

However, the fact that there is so much activity associated with Ka-band product development doesn’t necessarily mean that all this activity is a good idea.

Certainly that’s the view of Comsat Labs’ Evans. While supportive of many Ka-band applications, he warns that this spectrum isn’t just a higher frequency version of the C- or Ku-bands.

Chief among Evans’ concerns is Ka-band’s vulnerability to rain fade. He’s not impressed by satellite manufacturers and their claims of 99.5 percent availability for Ka-band. That’s because “at C-band we get typically 99.95 percent,” says Evans. This makes this older spectrum “almost as good as the telephone. There’s very little outage, and what there is tends to be caused by satellite problems rather than propagation interference.”

In contrast, Ka-band’s 99.5 percent availability rate works out to “40 hours a year” of downtime, says Evans. Moreover, this 99.5 percent is a global average. Actual outages are likely to be worse in the tropics, where there is a great deal of rain, as opposed to dry northern areas. This explains why he believes Ka-band may work for many applications, but not those like rural telephony and telemedicine, where being able to call out is a round-the-clock must.

Evans’ second concern about Ka-band VSATs is their price tag. “You’re forced to design the system in ways to combat rain fade, and that raises the cost,” he says.

In the case of Ka-band earth terminals, this means adding transistors to boost the output during rain fades. As well, the industry’s solution for this problem–having the earth station’s output automatically boosted anytime a signal dropout is detected–will also cost money: both in hardware and electricity. There’s just no way around it.

Finally, there’s the Iridium trap. Contrary to the movie Field of Dreams‚ just because you build it, doesn’t mean they will come. Certainly that’s a bitter billion-dollar lesson Motorola learned the hard way. Now the global voice service is getting a second chance as a new group of investors try to reignite the venture.

Could the same thing happen with Ka-band? “Yes,” says Evans. Although there are corporate and government uses for this spectrum, “few of us believe that a good consumer market exists for Ka-band,” he notes.

“The problem is very much like the problem that Iridium ran into: namely that yes, there’s a market today, [but only] if you can launch the system and produce a terminal at low-enough cost, and by low-enough cost‚ I mean less than $1,000,” he says. “But by the time any of these satellites are up and operating…the likelihood is that ADSL (asymmetric digital subscriber line) and cable modems will have reached a very large fraction of the potential market. Worse yet, they will be cheaper; almost everybody agrees that the terrestrial technologies will be cheaper than satellite.”

Defending The Dream

This is heady criticism, and it strikes at the very heart of Ka-band for VSAT applications. So what do the other players think?

When it comes to rain fade, Robert Bauer–the ACTS project manager at NASA’s Glenn Research Center–says Evans does have a point. But only in some cases.

For instance, if you use Ka-band to launch an Iridium-style service “where you have a small handset that uses a satellite link, then what he said might be valid,” says Bauer. “On the other hand, if you look at a DirecPC-style system where you’re using Ka-band to establish a wideband link…you could design a fixed terminal, which could be a little bit more robust.”

Meanwhile, when it comes to Ka-band’s unclear business case, “we agree with Evans’‚ position,” says Gilat’s VanBeber. That’s why “we are waiting for a while before switching our VSAT base over to Ka-band. We do think there’s a customer cost issue associated with this equipment: that while you may be able to offer higher bandwidth, the business case that goes along with it is not particularly attractive right now.”

However, Spaceway’s Cook has no doubt that Hughes can overcome all of Ka-band’s VSAT obstacles. “As far as Ka-band is concerned, we’ve obviously done a lot of work on what we think the propagation issues will be,” he says. “We (HNS) will be offering Spaceway network availability at least equivalent to our Direcway services today. We certainly expect to be signing service level agreements guaranteeing network availability well in excess of 99.5 percent. The bottom line is we absolutely expect to be able to deliver high levels of availability.”

As for cost? “We are very, very confident that we’re going to have a very effective end-user terminal which, frankly, is likely to be cheaper than the Ku-band VSAT systems that we’re selling today,” says Cook. Asked how this could be possible, when Ka-band terminals require more sophistication, he replies, “We know what we’re doing. We’re the major manufacturer of Ku-band systems today, [and] we are intensively involved in the development of the systems right now, and working with the subcomponent suppliers. So we’re very confident that we’ll be able to meet the cost targets.”

Finally, there’s the all-important issue of Ka-band’s actual marketability. Well, forget Evans’ doubts: according to Cook, “the business case for the system is extremely robust.

“This is not in any way like an Iridium system,” he adds. “It’s almost not worth comparing them, because they’re so different. Iridium was aimed at the mobile voice market, and because of the cost of implementing the LEO architecture, it was clearly not going to be able to compete with terrestrial services where terrestrial services existed. This narrowed Iridium’s market to regions of the world where there was no regional infrastructure–the underdeveloped regions of the world. In other words, regions of the world where there isn’t as much disposable income to spend on these sorts of systems.

“The second problem is the way Iridium managed its distribution, where it often was relying on organizations that had an investment in competing terrestrial infrastructure,” Cook says. Finally, Iridium’s last flaw was that “the LEO system technically had a number of problems.”

In contrast, Spaceway’s Ka-band technology will let it “compete aggressively against terrestrial systems on a price-to-price basis, and a price-to-performance,” says Cook. This doesn’t mean that Ka-band VSATs will win the entire market, but that it will get a good share of the high-speed broadband business. Cook also believes that Hughes’ existing VSAT customers will move onto Ka-band eventually, because they’ll be getting more bandwidth for about the same money.

As for Astrolink? Azevedo echoes Cook’s bullishness about Ka-band VSATs. “We will be an excellent value proposition for our customers,” he says, “in terms of being a one-stop shop that can provide them with a virtual private network and access to the Internet, together with a substantial number of business applications.”

Clearly, Azevedo has no doubts that Ka-band VSAT service is marketable. Or does he? That depends how you view Azevedo’s plan to launch Astrolink soon over existing satellites– “or even terrestrial technologies”–rather than waiting until its constellation is in place.

Whatever your opinion, this tactic will allow Astrolink to market-test its concept before spending billions on Ka-band satellites. Asked if this will protect Lockheed Martin from an Iridium-style fiasco, Celso Azevedo replies, “exactly.”

The Big Picture

Which brings us back to the big picture: is there really a market for Ka-band VSATs? Or are we seeing the kind of thinking that made the Iridium failure happen? At this point, there’s no clear answer to these questions.

On the one hand, the people preparing to launch Ka-band VSAT services are keenly aware of what went wrong with Iridium, and are even keener to avoid the same fate.

On the other, what really killed Iridium is exactly what worries Evans about Ka-band VSATs. When Iridium was proposed and financed, it was a good idea. The problem is that the market changed by the time its satellites were on station and ready to go. Had Iridium been able to launch within a few months of its conception, it would have stood a far better chance of success.

But then, that’s the satellite business. You just can’t rush these things off the ground, no matter how much you want to.

As a result, Astrolink’s test-it-now strategy seems to make sense. By seeing how many people actually want satellite broadband, Lockheed Martin will find out whether Ka-band VSATs have more commercial staying power than their satellite telephone counterparts. Certainly it’s a better–and cheaper–option then sending up satellites doomed to nothing more than re-enter the atmosphere.

This said, the sheer data carrying power of the Ka-band does make it a natural for VSATs. Someday. What remains to be seen is how, and when.

James Careless is a contributing writer to Via Satellite.

---