Internet Via Satellite: Ka-band and the Internet

by Peter J. Brown

Over the next 24 months, Ka-band–@29 GHz–will be moving rapidly from its status today as a somewhat minor league player in the satellite-based Internet realm to a more formidable role as a designated hitter. The possibility that Ka-band will contribute enormously to the evolution of the emerging broadband sector in the first decade of the 21st century represents the next significant chapter in satellite history.

For consumers in particular, having two-way access to affordable broadband services, independent of the availability of terrestrial infrastructure such as high-speed cable plant or telco-based DSL, is as exciting today as the launch of satellite-based multi-channel Ku-band DBS services was in the early-to-mid 1990s.

The developmental work done in Europe, Japan and the United States with a variety of Ka-band platforms is now giving way to commercial ventures, starting with a pair of Ku/Ka- band hybrid satellites launched recently in Europe by Betzdorf, Luxembourg-based SES Multimedia S.A.–using Astra 1H–and in Asia where Seoul-based Korea Telecom has deployed Koreasat 3. Denver-based Wildblue Communications Inc.–formerly known as ISky–appears to be lining up as the Ka-band service market leader in North America with its own Loral- built Wildblue 1 satellite going up in late 2001. This will be followed by Wildblue 2 in late 2002, which consists of 90 percent of the 45 Ka-band spotbeams aboard Telesat Canada’s Anik F2, although this arrangement is still subject to FCC approval.

“Ka-band and the new technologies, through the use of spotbeams, higher power and increased spectrum, will bring lower air time and lower terminal costs,” says Paul Bush, Telesat Canada’s vice president of corporate development. “The challenge with a new technology is that it is a new technology. There is a lot of work to do on gateways and terminals, standards are still evolving, and applications are still germinating.

“I am, however, confident that fiber and DSL will never blanket the country like satellite. And 30 to 40 percent of the population is a big number for us to have as a market,” Bush adds. “I also believe that satellite–like it has done in the broadcasting sector–will revolutionize the Internet and how we use it as the Internet is forced to deal with efficiencies, and the reality of streaming video and audio.”

A Whole Different Ball Game

Jeff Leddy, vice president of strategic corporate planning at Norcross, GA-based EMS Technologies Inc., emphasizes the shift to higher frequency satellite communications is a challenge, but it is a separate matter entirely from how the technology is implemented.

“Business plans are not just dependent on satellite technology,” Leddy says. “You often encounter situations where satellite system operators put a marvelous technological asset in space, while what is taking place on the ground results in two completely different realities.

“We are developing Ka-band technology for spacecraft such as Astrolink and Anik F2, as well as Broadband Interactive (BBI) ground terminals for SES Astra,” Leddy adds.

Leddy points out that there will be a strong demand for solid state amplifiers in the 1/4 to 2-3 watt range at attractive price points and volumes. Could lack of an adequate supply have an adverse impact over time? Leddy did not go quite that far.

Leddy credits SES Astra and Koreasat for setting the Ka-band market in motion with their hybrid Ku-/Ka-band satellites. Still, the issue of frequency re-use goes virtually untouched thus far, especially with the Astra 1H satellite.

“Frequency re-use is present in most of the proposed Ka-band architectures. With respect to Astra 1H and the BBI interactive terminals, there are a very limited number of beams on the uplink side, and the benefits of frequency re-use are not realized, while more bandwidth is available,” Leddy says.

“We will not see any substantial frequency re-use by SES Astra in Europe until Astra 1K goes up in 2001 with its 16 spotbeams, and in North America with the launch of Wildblue and Anik F2 in late 2002,” he says.

SES Astra has formed a 50/50 joint venture with IBeam Broadcasting in London that will allow for IBeam’s Maxcaster servers to facilitate large file delivery and Webcasting to the “edge” of the Internet. With respect to two-way satellite services, Robert Feierbach, business director at SES Multimedia’s broadband interactive group, indicates that by using multi-frequency time division multiple access (MF-TDMA) for frequency hopping, the single Ka-band transponder on Astra 1H could theoretically accommodate all the data traffic on the return channel generated by 500,000 to 600,00 typical consumer broadband end-users. This assumes that the entire transponder was allocated to bursty consumer traffic.

The BBI fully redundant hub governs the allocation of traffic between all users on an on-demand basis.

“Given the nature of this shared hub environment, we are carefully evaluating our overall compatibility with European cable operators who are rolling out interactive services, as an early entry consumer model for BBI,” says Feierbach. He adds that for small cable operators in particular who want to provide a low-cost consumer solution using a DOCSIS modem as customer premises equipment, it is possible that the cost of a 2 Mbps BBI terminal could be shared and combined with a simple cable modem router.

With IP multicasting just becoming viable, Ku-band satellite-based solutions are drawing considerable attention well in advance of any large-scale Ka-band deployments.

“We put all the pieces together finally in the last year and a half. Now, these advances in IP multicasting have convinced a lot of people that if you want to deliver large amounts of IP content to multiple sites, satellite is the way to go,” says Toby Farrand, president and CEO of Milpitas, CA-based Broadlogic Network Technologies. “We are very active in Ku-band, providing our customers such as Starband Communications with receiver units.”

“With Ka-band, the driver applications are going to be quite different than was the case for Ku-band. Ku-band is already becoming quite crowded, and IP will consume a lot of bandwidth. The rise of IP data delivered over satellite will outstrip the Ku-band bandwidth available. Pioneer Consulting, for example, indicates that the trend points to a phenomenal growth in streaming media, to the point that 70 or 80 percent of all IP traffic in the coming decade will be streaming content,” he says.

Farrand sees a range of Ka-band satellite projects proceeding, and notes that “some will be riskier than others.” Astrolink, for example, is counting on advanced Ka-band satellites with onboard processing that hold the promise of better network performance, but at the risk of greater complexity. There are lots of wild cards too, including the need for whole new subscriber management schemes, especially as consumer gateways start to proliferate, and as inbound satellite-delivered feeds flow not just directly to PCs, but also to PCs that have been reconfigured into hub routers. Regardless, the winners will still be the companies that are best able to consistently deliver a rich, compelling user experience.

“As different modulation schemes such as 8PSK and 16 QAM are tied to data streams using 5/6 or 7/8 encoding, for example, not only will we see increases in throughput from double-digit to low triple-digit Mbps, but we are also seeing changes in routers from 100BaseT on up to Gigabit Ethernet and ATM,” Farrand says. “Higher power satellites and their ability to close the link budget represent the only way to keep stride with all these related developments in terms of providing an affordable streaming content broadcast service.”

Farrand believes a key challenge for many companies at the dawn of the Ka-band era will be staying tightly focused. Broadlogic, for example, intends to limit its data over satellite activities to the receive side of the end user equipment market despite the obvious attraction of the headend systems market.

“As ‘co-opetition’ takes hold, it is becoming so easy to lose track of what you do best,” Farrand says as he watches encapsulation providers dabble in terminals, among other things. “This marketplace is very unforgiving.”

Aiming For A Single Standard

Despite all this activity, Ka-band remains merely a hypothetical force in the broadband marketplace. By early 2003, we will have a much better understanding as to how far the satellite industry can pitch this ball.

“How are we going to ramp this up and grow it? What needs to be done now to promote growth and mass market adoption? These are the two fundamental questions today,” says Gordon Deans, vice president of Norsat Broadband Networks in Burnaby, British Columbia.

Norsat is playing a key role in SES Astra’s BBI system, providing layered DVB-based broadband services throughout Europe, including Astranet’s multicast file delivery and streaming media services that are available at speeds of up to 6.5 Mbps.

In addition to its recent shipping of Ka-band terminal prototypes to Korea Telecom for testing, Norsat is teaming up with Norcross, GA-based EMS Technologies Inc. Norsat is providing outdoor units (ODUs) while EMS is providing the indoor unit, and together these constitute the SES Astra BBI satellite interactive terminal (SIT).

“Component availability, and the quality of these components, is not an issue. What we need to address is what it is going to take to carry out large-scale deployments, and this includes attracting the necessary investments for the establishment of space and ground segments,” says Deans.

He is a strong proponent of a common standard. One of Deans’ biggest concerns is that the satellite industry will roll out Ka-band services as an assortment of different offerings based on competing, proprietary standards. DVB-RCS offers the industry the opportunity to embrace a single return path standard. Thus far, DVB-RCS has not been embraced with much enthusiasm outside of Europe.

“To achieve economies of scale similar to alternative technologies, it is important that we adopt a common standard in the Ka-band market, and that the effort is made now to build a high-growth infrastructure,” Deans says. “We will not be given the opportunity to launch Ka-band twice.”

Will the failure of the satellite industry to identify and adopt a single acceptable standard such as DVB-RCS really handicap the industry as it attempts to position Ka-band as a viable broadband contender? As you can see from Deans’ comments above, there are strong opinions about what is at stake here, and how essential standardization is as a catalyst for rapid, large-scale deployments.

Now Up To Bat: The Sec

Although the Federal Communications Commission (FCC) and the ITU are both organizations where important decisions are being made that bear directly on the future of Ka-band- based ventures, there is another agency that is now taking center stage: the U.S. Securities and Exchange Commission (SEC). It is here that companies such as Wildblue and Bellevue, WA-based Teledesic Corp., which is merging with New ICO via a transitional entity in Kirkland, WA, called ICO-Teledesic Global Ltd. (ITGL), are making important disclosures. ITGL will cease to exist once the New ICO merger is completed, and a Teledesic spokesperson indicated that this should be completed by the end of 2000. ITGL has filed merger-related documents, while Wildblue is filing securities-related documents in late 2000 that help highlight the risks and rewards in both these and other similar Ka-band ventures, among other things.

The risks in terms of increasing competition in the sky, and on the ground, along with a variety of technological and even regulatory snags are numerous, and not easily dismissed.

For example, consider this statement concerning capacity by ITGL on p. 18 of the S-4 submitted to the SEC. It is important to remember that the impact of latency here is one of the key reasons why Teledesic locked on to the concept of a LEO constellation consisting of 288 satellites from the beginning. Teledesic’s list of backers include Craig McCaw, who acquired ICO last May for $1.2 billion after ICO had filed for Chapter 11 protection in late 1999, and Microsoft founder Bill Gates, along with The Boeing Co., Motorola, Saudi Prince Alwaleed Bin Talal, and The Abu Dhabi Investment Co.

“The spectrum and satellite infrastructure characteristics of our networks create inherent limitations that prevent capacity expansion. Failure to achieve a commercially viable level of capacity would adversely affect us,” states this S-4, which labels frequency coordination as a key concern before proceeding to explore the impact of usage patterns. This impact has loomed over the consumer side of the satellite-based Internet services arena since the launch of the hybrid platform known as DirecPC by Germantown, MD-based Hughes Network Systems in the mid-1990s.

“Actual usage patterns may differ from our expectations. We could experience unexpected usage patterns that exceed the capacity of the networks through one or several satellites. If we face significant capacity issues resulting from inadequate spectrum availability, we may face significant regulatory hurdles to increase our spectrum rights.”

Teledesic was in a quiet period as this article was being prepared, and this prevented Teledesic executives from discussing anything other than the big picture. It is planning to emerge in two separate stages, beginning with the launch of New ICO’s Iconet-based mobile satellite services in 2003, followed a year or so later by the satellite-based broadband services of Teledesic. However, there is a defined migration path for any interested ICO customers from lower data rates to Teledesic’s high bandwidth-based service grid, as well as a plan for Teledesic to use the Iconet ground infrastructure. New ICO announced in September that Boeing–now the parent company of Hughes Space and Communications Co. (HSC)–would build three additional 601 series satellites, while making modifications to the eleven 601 series spacecraft already in production.

This notion of a growing synergy on the basis of shared ground infrastructure, among other things, between ICO and Teledesic surfaces time and time again in ITGL’s S-4. For example, on p.116, a broadband repeater capability involving the installation of an external satellite antenna is explored.

“Thus, any cell phone, personal digital assistant, laptop or other mobile device equipped with a modem could access the Internet or Intranet at high speed and inside buildings at an expected range of 100 meters. This product would be a natural evolution of some of the products that are expected to be offered by New ICO. Teledesic views this product as an ideal solution for deploying third generation-type cellular services in remote parts of the world.”

The Challenges Of TT&C

Synergy aside, a lot of minute details need to be addressed. Consider, for example, how these Ka-band satellites will actually be deployed. In Wildblue’s S-1 SEC filing, one encounters yet another challenge. This time it involves the uncertainty surrounding certain aspects of the telemetry, tracking and control (TT&C) of Ka-band satellites in the early phases of their flights.

“FCC rules anticipate that a satellite operator will use spectrum at the edges of its authorized band to conduct TT&C operations. However, because very few Ka-band satellites have yet been launched, many parts of the world have no facilities capable of communicating with a satellite in the Ka-band,” the Wildblue S-1 states. “This presents a particular problem during the launch and initial transfer orbit stages of our satellites’ deployment, during which we will need to perform TT&C over a wide portion of the international orbital arc, and in case of an emergency, when we may need to draw upon TT&C capabilities from many locations in order to establish contact with our satellite.”

Wildblue goes on to describe how it will conduct these satellite TT&C operations in the much more established C-band frequencies.

“However, because this would be a non-conforming use of spectrum under the FCC’s rules, we are not yet authorized to use these frequencies for this purpose. In order to receive the necessary authorization, we must demonstrate either that our TT&C operations will not interfere with other conforming operations in the band or that we have successfully coordinated our TT&C operations with all affected operators in the band. We have not yet begun coordination nor filed the required application,” Wildblue states.

Wildblue points out that, “although the FCC has granted authorization for similar non-conforming uses in the past, we may not be able to make an appropriate demonstration and secure the required license modification.” Until this gets resolved, Wildblue is stuck in a holding pattern, as the following sentence indicates.

“Failure to obtain this authorization for the launch and transfer orbit phase would preclude us from launching our satellites, while failure to obtain it for emergency operations would leave us vulnerable in case of a catastrophic failure of our satellites’ normal TT&C subsystems,” states Wildblue, adding that if the company is “unable to obtain authorization to use alternative spectrum, we will have to use some of our Ka-band spectrum for TT&C operations, which could increase the cost of these operations due to the general lack of Ka-band facilities.”

A Dynamic Environment

TT&C considerations aside, the emergence of various Ka-band business models hinges on the successful implementation of a satellite-based solution that is able to meet the demands of what will be a very dynamic environment. There is a sense that the challenges posed by this dynamic environment have been both identified and overcome in earlier Ka- band trials. For example, consider this statement which appears in Wildblue’s S-1 filing at the SEC.

“Although Ka-band spotbeam satellites are relatively new, Ka-band technology has been successfully tested by NASA’s Advanced Communications Technology Satellite (ACTS) program, which launched a Ka-band satellite in September 1993. We believe the ACTS program, which ended last year, has proven the viability of a two-way spotbeam system using Ka-band frequencies, and it has provided a large public database of Ka-band signal propagation information over North America,” states Wildblue’s filing on p.41.

David Beering, a consultant at West Chicago-based Infinite Global Infrastructures LLC (IGI), is intimately familiar with ACTS, and he is not prepared to offer ACTS up as the ideal model for the next generation of Ka-band platforms.

“ACTS had to be manually reconfigured several times a day. Besides addressing the best architecture for aggregating traffic, people have to look at this new bi-directional data networking capability as a new phenomenon. Readers must be mindful of the fact that ACTS was a dedicated platform designed to function in a static experimental configuration, and it did not operate in the dynamic, shared network environment which is so widely discussed today,” Beering says.

ACTS has been a success as far as validating the potential for Ka-band platforms; however, as it approaches the end of its lifetime, suitable off ramps for many of the ACTS users need to be located relatively soon. According to Beering, Telesat may well hold the key for this effort in North America.

“What Ka-band requires in terms of automation, and uplink power control where hundreds of thousands of terminals are involved, was not part of the ACTS equation,” Beering says.

“People are scaling back from full Ka-band implementations, and Ku-/Ka- hybrids are attracting more attention, thanks in part to advancements in networking, access to numerous applications over IP, the ability to achieve quality of service via dedicated space segment–and not just ATM but also Multi-Protocol Label Switching (MPLS), for example, which while not now implemented is available–and other breakthroughs such as the assorted reliable multicast protocols,” Beering adds.

Taking A Swing At The Fcc

As the FCC proceeds to sort out the shape and look of the Ka-band landscape, it faces continued criticism from first and second round Ka-band license applicants. FCC actions and inaction alike are seen as hindering rather than assisting the launch of Ka-band services in the United States.

The FCC is being blasted for what amounted to “poor public policy as well as a serious injustice,” according to EMS Technologies, due to the FCC decision to revoke the Ka-band license of Netsat 28 in mid-2000 after it had been acquired by EMS Technologies in late 1999. In addition, it appears that at least one of the second round applicants, Pegasus Communications Corp. in Bala Cynwyd, PA, is not at all happy with what is unfolding at the FCC as well.

“With the exception of Netsat 28 and a couple of others, the FCC has been very slow to revoke licenses. In early 1997, the FCC waived its two-slots-per-applicant rule and gave some companies 20 or more slots all at once. But the FCC said it would revoke those licenses if the incumbents did not launch satellites quickly,” says John Hane, the Washington, DC-based vice president for business development at Pegasus Development Corp.

In addition to Pegasus Development Corp., other second round applicants identified by Hane include DirectComm, CAI Wireless–a Worldcom subsidiary–Lockheed Martin, which already has licenses for 10 satellites at five slots in the first round, Hughes, which has more than 20 satellites in the first round, Panamsat, which like Netsat 28 had two of its first round licenses revoked for failure to meet construction milestones, Loral, Motorola, Ka-Star, TRW, Celsat, and Pacific Century Group.

“The idea was that strict enforcement of the milestones would ensure sufficient slots for the second round applications, which were filed just after the first round of licenses was issued,” Hane adds. “Instead, second round applicants have been coping with a dearth of licenses, while the incumbents–companies like Hughes, Loral, GE Americom, Motorola and Panamsat–have no incentive to actually build Ka-band satellites or constellations.”

Hane compares this state of affairs to what might have happened 30 years ago if, on a hypothetical basis, the U.S. government had granted exclusive rights to build microcomputers to IBM and Digital Equipment Corp., and then told Apple Computer four years later that it was too late.

“Imagine where we would be now. What is going on at the FCC is a travesty. Normally, almost four years of holding a license without building would trigger an automatic loss of the license in question,” Hane says. “Notably, 43 months after licensing, the only consumer-centric Ka-band system funded and building is Wildblue, a new entrant with no Ku-band business to protect. Our unserved rural population deserves more new entrants–Pegasus and others–building satellites, and fewer FSS incumbents warehousing precious orbital slots.”

Hane believes the initial FCC licensing plan for Ka-band licenses was done on the basis that companies could take all they wanted, “but they had to eat all that they took.” With the broader yet critical ITU timetable looming as well, which extends to 2004, according to Hane, the Ka-band market is being subjected to a handicapping system that only serves the interests of the larger entrenched satellite operators. And this is what second round applicants are colliding with despite more than two years of negotiations that were aimed at creating a rational allocation of the remaining Ka-band slots.

“Ka-band is a fundamentally different equation altogether. It facilitates a shift from a broadcast market to an access technology-driven market, and it can put the power of the satellite directly in the hands of the end users at very low cost. At best, the incumbent satellite operators who hold all of the licenses do not understand this business and are afraid to invest. At worst, they hope to stall the launch of Ka-band services because Ka-band amounts to a real threat to their existing business models,” Hane says.

Promise And Potential

Consumer awareness of the promise and potential of satellite-delivered Internet services is gaining considerable momentum. This trend is fueled by the recent launch of two-way Ku-band Internet access services by Starband Communications in the United States–Hughes Network Systems’ DirecPC had not yet launched their two-way version at press time–and also because of companies like Luxembourg-based Europe Online Networks S.A. Europe Online launched in September 1999 and is reporting spectacular results, offering a product which uses a hybrid Ku-band satellite/terrestrial return path.

Founder and founding president Candace Johnson appears to be remarkably upbeat about what is unfolding. Europe Online operates a hybrid celestial/terrestrial network with combined multicast, unicast, and push-pull capabilities. This company has also installed a fiber optic Internet backbone and terrestrial access network to complement its satellite network. At its uplink site, Europe Online is using products from Harmonic Data Systems, Inktomi, Cisco Systems, Sun Microsystems, and Skystream.

Europe Online has Technisat, Hauppauge Digital, Harmonic, and Broadlogic as providers of its PCI cards, while SCM supplies a USB box, and Strong and Siemens-Fijitsu serve as vendors for Europe Online’s TV set-top box. Techno Trend is the vendor for its cable PC card.

“Europe Online continues to work with all of these vendors, but we have now personalized and branded the cards under the Europe Online brand. You can now buy a subscription for $17.00 per subscription per month, and get the card free with the subscription,” Johnson says.

“We have not only signed content deals with ZD Net, EuroSport, CNBC and Travel, to name a few, but we have also signed content deals with Sony and Bertelsmann Music Group as well as having our own Europe Online Digi Studio and mass consumer call center,” Johnson adds.

Johnson reports that Europe Online raised over $60 million in early November, and has recruited a former high-ranking executive at The Walt Disney Co., Dennis Hightower, to serve as CEO.

“We are pushing content and have a lot of it…. Also, we actually offer the digital television, radio and broadband Internet all in one. It is a very converged experience, and you can go between the two and have them all at once–digital television, streaming media, Internet and downloads,” Johnson says.

No wonder the satellite industry is so excited about the possibility that a concerted Ka-band drive will be getting underway in the near future.

Implementing Ka-Band will require some delicate footwork. In part two of this article, we will examine emerging platforms and business models which must be able to adjust to a wide range of variables including increasing terrestrial wired and wireless competition, and, the proliferation of distributed networks in key markets.

Peter J. Brown is Via Satellite’s Senior Multimedia Editor. He lives on Mount Desert Island, ME.