Virtual, Wireless Manna: Interactive Multimedia Applications and the Broadband Satellite Industry

Elias G. CARAYANNIS <caraye@research.circ.gwu.edu>
George Washington University
USA

Jeffrey ALEXANDER <jeffalex@wcore.com>
Washington CORE / George Washington University
USA

Abstract

The emerging satellite communication constellations differ substantially from prior generations of systems. For years, satellite systems have offered broadcast and interactive services, but with major limitations. The new generation of services promise to overcome the traditional barriers, such as terminal weight and size, equipment and service cost, and degree of mobility. Instead, the new systems will provide completely new capabilities which expand the flexibility for the user and the range of potential applications. The development of broadband satellite systems is the first step in the creation of a market for multimedia satellite services. Such services are composed of three layers of technology, all of which are undergoing constant development and extension. The first layer is the delivery technology, which sets the fundamental technical limitations of further service development. The second layer consists of multiple service platforms. These platforms create communications environments which define how end-users interact over the system. The third layer is the set of applications available to the end-user. These applications and their corresponding value-added services are software environments designed to enable end-users to accomplish a specific task. This paper discusses the likely evolution of multimedia applications delivered by satellite using this three-layer framework.

Keywords: Wireless Multimedia, Next Generation Internet Technologies, Co-opetition

Introduction

The global information infrastructure (GII) is undergoing tremendous change as demand for broadband communications services continues to outstrip supply. While a number of new ventures will attempt to fulfill that demand by greatly increasing transoceanic fiber network capacity, there is substantial doubt that wireline facilities alone will be both adequate and technically and financially feasible for providing the total bandwidth required to sustain the GII. Satellite networks are already playing a role in supporting the bandwidth demands of the Internet as a backbone system, through such enterprises as SkyCache. Other satellite service providers such as Comsat provide bandwidth on demand to major landline carriers to provide redundancy for wireline networks or to supplement available bandwidth. However, an emerging generation of satellite systems now promise to act as a primary carrier of interactive broadband services, putting new demands both on satellite technology and enabling the provision of new applications.

Current satellite communications systems are optimized to support high-demand, mass market services such as VSAT (very small aperture terminal) communications (used primarily by corporations and government), live video broadcasting, direct-to-home or direct-broadcast-satellite (DBS) television, and long-haul voice telephony. The ground station equipment and terminals tend to be large, fixed, and very expensive. Therefore, in almost all cases (except DBS service, which entered widespread use in the United States in approximately 1997), the major users of satellite services have been institutional users, including large businesses, government agencies, and telecommunications carriers.

In place of the current generation of systems, private companies are now launching new satellite communication systems which overcome these barriers. This new generation of systems is notable primarily for its use of low-earth-orbit (LEO) architectures, instead of the conventional geosynchronous (GEO) orbit configuration. Since the number of GEO orbit slots is limited, and signal latency to and from the GEO orbit is significant, LEO systems provide superior performance in terms of user mobility, signal quality and latency, terminal size, and availability of orbital slots. Other systems use mid-earth-orbit (MEO) architectures, or in some cases hybrid GEO-LEO and MEO-LEO configurations.

Overview of market assessment issues

The recent record of market assessment and forecasting in the satellite industry is mixed at best. Limiting the review to new services aimed at the general business and consumer markets, some have succeeded beyond the expectations of analysts, such as direct broadcast satellite (DBS). After its U.S. launch, DBS service gained the record for the fastest rate of adoption of any consumer electronics product (until that record was recently surpassed by digital video disk players). On the other hand, the widely anticipated market for Global Mobile Personal Communication Systems (GMPCS), provided by firms such as Iridium and ICO, has failed to materialize. As with any advanced communications service, predicting the adoption of new satellite services by individuals not a very easy endeavor.

Assessing the market success potential of broadband satellite systems requires the consideration of a broad range of factors. Only some of those factors are within the span of control of satellite operators. Therefore, operators will have to make choices as to which factors they will address in their business strategy. In the domain of technical factors, as an illustration, decisions on system technology and design allows operators to make trade-offs among various capabilities; for example, choosing a MEO system design over a LEO system reduces complexity but increases latency and other disadvantages.

The costs of a poor market assessment are enormous. The recent experience of the Iridium satellite constellation provides some insight into those costs. Executives at Iridium have admitted publicly to some poor choices in market assessment and strategy. For example, the company targeted its services towards global business users. The added functionality of Iridium did not justify the cost of the service; furthermore, global roaming provided a very small addressable market for Iridium, as global business users account for only a tiny fraction of total mobile telephone usage. As Iridium was unable to meet its target dates for subscriber levels, the company was unable to meet the covenants on its debt. Iridium's major creditors forced the company to declare bankruptcy and enter Chapter 11 restructuring proceedings.

There is now some fear that the bankruptcy of Iridium has "poisoned the well" for future satellite systems. Most broadband satellite systems require billions of dollars in capital for satellite construction, ground station networks, launch costs, and other expenses. Due to their design, LEO and MEO systems must have enough satellites in orbit to provide uninterrupted global coverage before they can enter service -- this means that the up-front investment required is substantial. While most of the broadband satellite systems proposed so far have secured sufficient capital to start operations, there is the fear that if any of these systems runs into unexpected expenses (due to a high rate of launch failures, for example), investors will not be willing to give them the money to start business.

Given these very high stakes, a careful consideration of the exact position of broadband satellite systems in the competitive broadband communications market is needed. This entails the assessment of the strengths and weaknesses inherent in broadband satellite service, and how those characteristics affect the threats and opportunities presented by the market environment (see table 2).

Competitive advantage for broadband satellite

Broadband satellite systems are entering a very crowded market for telecommunication services. With competition from broadband wireline and terrestrial wireless technologies, satellite communications is faced with a tremendous challenge to market success. Still, these systems promise a number of features without a comparable match among other technologies:

• Global coverage. LEO and MEO systems in particular can deliver service to any location regardless of terrain or existing infrastructure.
• Instant deployment. With satellite systems, a communications service can be deployed globally with minimal time and cost for set-up.
• Bandwidth. Broadband satellite systems will offer "bandwidth-on-demand," where users receive and pay for capacity only when it is used.
• Cost. Broadband satellite systems expect that their services will be at or slightly below the costs of wireline services in developed countries, but with more features and convenience than wireline.
• Uniformity. Multinational organizations can replace multiple regional carriers with a single point of contact offering a unified, global menu of services.

There are a number of potential barriers to the success of individual systems and the broadband satellite industry as a whole. These barriers can be classified as technical, financial, and regulatory factors.

Technical factors

Management of latency. Most satellite systems, particularly GEO systems, suffer from significant and variable latency. While this is not a problem for asynchronous packet-data services, such as e-mail, it will be an issue in using packet data networks to support synchronous communications such as voice and video broadcast service. Proposed satellite systems have developed proprietary solutions to the latency problem, most based on "IP spoofing" (where IP packet transmissions generate a false acknowledgement, so that the transmitter does not attempt to resend the packet before it has actually arrived at the destination). Unfortunately, these solutions lock customers into specific hardware for each satellite system, increasing switching costs relative to other broadband technologies. Also, spoofing may not function in conjunction with standard network security measures.

Quality of service. In addition to latency, the quality of all types of packet services may be degraded in satellite systems by factors such as rain attenuation, signal strength, and system failures. Therefore, the broadband systems will require sophisticated error correction techniques, but even these may not guarantee a level of service equal to competing technologies. Satellites do have one advantage in quality of service. Most terrestrial Internet services depend on interconnection between networks to deliver transmissions. This means that a high-quality carrier may be hurt if its transmissions must travel over a low-quality carrier's network to reach their destination. Satellite systems provide total end-to-end connectivity over a single network, so that quality of service is more predictable than for competing technologies.

Complexity. For the LEO systems (Skybridge, Spaceway, and Teledesic among others), the large number of satellites and the frequency of handoffs between satellites to maintain a communications link pose a tremendous challenge in managing the complexity of the communications system. For example, Teledesic is based on a system design requiring 288 satellites in orbit simultaneously (a reduction from the original design, which called for 840 satellites). While the "on-board processing" approach of these systems offers advantages over the "bent-pipe" architecture of conventional systems, this also requires extremely sophisticated network management software that could be subject to failure.

Financial factors

Availability of capital. All of the proposed systems are tapping the financial resources of major institutional investors and the global public capital markets. Since almost every market forecast predicts that no more than three or four broadband satellite systems will be commercially viable, there are few sources of capital which can provide the amounts of investment needed to launch these systems. Therefore, if an operator fails to convince a few investors to back its project, it has few options for recourse.

Investor confidence. Capital availability may be further limited by adverse developments. In the narrowband segment, capital has almost dried up following the failure of several launches of satellites for the Globalstar system and the financial distress experienced by Iridium. Similar technical or marketing setbacks in the broadband sector could cause investors to back out of the proposed systems, leaving some or all with inadequate resources to complete their systems.

Predictability of capital costs. All of the systems may run into major cost overruns during deployment and implementation that may exceed available resources. Sources of potential overruns include delays due to launch failures, development costs for handsets, costs for ground station development and construction, and delays due to regulatory factors (see below).

Regulatory factors

Spectrum allocation. A major barrier in the Ka band was overcome when the World Administrative Radio Conference (WARC) allocated global bandwidth to systems operating in this band. Systems which utilize spectrum outside the Ka band face the possibility that not every country will allow them to operate in their desired spectrum band, meaning that they will not be able to achieve full global coverage.

Global telecom regulation and competition policy. Broadband satellite systems will compete directly with landline and terrestrial wireless communications carriers, many of which are government-supported or are in fact PTOs (posts and telecommunications organizations) owned and operated by governments. This may mean that national governments will limit the ability of the broadband systems to market and provision their services in those countries. The World Trade Organization has facilitated an agreement to liberalize international trade in Basic Telecommunications Services (BTS), which is essentially voice telephony. However, trade barriers to data communications (such as a "bit tax" once proposed by EU Commissioner Edith Cresson to tariff digital content imported into Europe) are not covered under these talks, and could be used to reduce the penetration of multimedia services in some countries.

Threats and opportunities in broadband communications

Given their capabilities, broadband satellite systems are best positioned to offer the services addressing four primary markets.

• Wholesale carrier services, providing high-capacity, worldwide carriage of communications traffic over long distances. This service is generally purchased by telecommunications carriers to establish global connectivity and provide redundancy for existing networks.
• Global corporate networking, providing both raw carriage and value-added services supporting data, voice and video communications for multinational corporations.
• Institutional users, providing high-bandwidth services to support public services distributed to large constituencies, such as education, healthcare, and welfare assistance.
• Broadband local loop, providing high-speed data and other services (especially Internet access) direct to homes and offices.

In each of these markets, satellite carriers must face potential competitors. In global wholesale service, Project Oxygen promises to deliver the same services as satellites using a global fiber optic network. In corporate services, satellite systems must compete with alliances among established players in the telecom market, such as the partnership between AT&T and British Telecom. In the broadband local loop market, one serious problem is that the potential competitors -- mostly wireline carriers -- are also the primary customers of the satellite industry's wholesale services. Some potential market-based threats facing these systems include:

• Elasticity of demand. The broadband satellite industry's growth is heavily reliant on very optimistic forecasts of demand for bandwidth across the world. While current growth rates for Internet usage seem to support these forecasts, the reality could be quite different. Also, it is not clear what prices users are willing to pay for broadband data access, which could also limit the viability of many systems.
• Competing technologies. There are several alternate technologies for providing broadband data service, including wireline technologies (cable TV data systems, digital subscriber line, fiber-to-the-home) and wireless technologies (fixed microwave, wireless cable, third-generation mobile telephony, and emerging wideband services). Depending on the cost and availability of the infrastructure for these technologies, the competition facing broadband satellite services could be very fierce in particular markets.
• Interconnection. To provide wholesale services, and to satisfy the preferences of some end-users, satellite networks must interconnect with terrestrial communication networks. The terrestrial carriers could tie interconnection agreements to other demands which could limit the ability of satellite systems to compete; for example, making interconnection contingent upon the agreement of the satellite operator to grant the landline carrier an exclusive distribution agreement in a particular market.
• Standardization. Delivery of specific interactive multimedia applications requires that the carrier support the various standards used to construct those applications. At a minimum, all carriers now must offer IP service to be serious contenders in broadband communications. But more sophisticated applications, such as virtual reality, will be based upon a dazzling array of standards covering quality of service, compression, data formats, and transmission protocols. In addition, industry groups are developing their own standards. For example, the U.S. automotive industry has created the Automotive Network Exchange (ANX), which now certifies which carriers provide IP services that meet the needs of the industry. ANX is being extended to include European and Japanese manufacturers. Other global industry-specific standards are emerging in chemicals, logistics, and electronics. Satellite providers will need to track all of these standards developments and ensure that they can, in fact, support the emerging dominant standards, or they will be shut out of important end-user markets.

Summary of challenges for broadband satellite market assessments

Market assessments for the adoption of advanced telecommunications services are complicated by a number of factors. First, network effects lead to the phenomenon of "tipping." Adoption of a communication technology is often very limited until some "critical mass" of users is reached, at which point adoption accelerates rapidly. This phenomenon is a function of the observation attributed to Robert Metcalfe, inventor of the Ethernet protocol, that the value of a network is a function of its number of connections. Some critics of technology forecasting argue that since the exact critical mass is difficult to predict with certainty for each technology, forecasting for new communications technologies is ultimately impossible (see Lucky, 1999, and Coates, 1999).

Second, technology innovators often do not realize the flaws in their own perceptions of the market. Innovators often fail to foresee secondary or downstream applications and consequences that arise after a technology achieves greater sophistication and widespread adoption. Also, there is a tendency to associate a particular technology with an emerging need, when in fact many alternative technologies may exist (Coates, 1998).

Third, there is the problem of trying to apply past experience to future opportunities. The danger in this is that the circumstances surrounding the introduction of a new technology may not exactly match those of previous technologies. For example, Iridium estimated the likely adoption of Global Mobile Personal Communication Services by estimating the current "addressable market" for global roaming mobile communications, and applying the growth rates of past terrestrial mobile communications to that number. In fact, GMPCS did not achieve those growth rates in 1998 and 1999, and in any event Iridium was unable to capture much of whatever demand did exist for GMPCS (Mertiens, 1997).

Various forecasting methodologies are intended to overcome these limitations, such as Delphi approaches or sophisticated quantitative models (Stordahl, 1997 and Mohamed and Lagacherie, 1999). Forecasters should keep in mind, however, that the purpose of a market assessment is to understand the dynamics of the market, not to calculate an exact estimate of future subscribership (Hewett, 1997). This means that satellite operator strategies should be based not on exact quantitative forecasts, but instead on investment decisions that are likely to increase adoption across a wide range of market scenarios.

Strategies for broadband satellite operators: a co-opetitive analysis

These industry and market conditions create a very complex competitive landscape for satellite providers, making it difficult for those providers to decide how they will deal with the many players involved in broadband communications. The framework developed by business researchers Adam Brandenberg and Barry Nalebuff in their book Co-opetition is a useful tool for mapping this landscape. This framework, called the "value net," classifies the players in a company's market into four categories: competitors, complementors, suppliers and customers (see Figure 1). For the broadband satellite industry, these categories might be configured as follows:

• Competitors: Other providers of global and local broadband services, including international carriers (especially Concert, the AT&T-British Telecom alliance), domestic and regional telephone companies (France Telecom, Deutsche Telecom, Cable & Wireless), cable TV operators (AT&T, Time Warner), wireless carriers, and other satellite operators.
• Complementors: Firms whose products and services will increase demand for broadband satellite services, including content providers (movie studios, television programmers, music publishers, entertainment companies), computer and semiconductor manufacturers (e.g., Intel), software developers (game companies such as Id Software), standards bodies (especially the International Telecommunications Union) and systems integration firms.
• Suppliers: Firms who provide goods and services used by broadband satellite providers, including ground station and transmission equipment manufacturers, handset vendors, software companies, launch services, and content providers.
• Customers: The key end-users of broadband satellite service, namely carriers, corporations, institutional users, and consumers.

Figure 1: The Value Net
Source: Brandenburger and Nalebuff (1996)

One key feature of this framework is that certain players can occupy multiple categories. For example, terrestrial communications carriers can be customers of the satellite operators' wholesale services, but competitors in corporate services and broadband local loop. Content providers are suppliers in one sense, and they may sell content directly to end-users through the operators' satellite systems, but complementors in that the general proliferation of broadband content will help attract customers to broadband satellite services.

Using this framework, satellite operators can begin to formulate key strategies that will help them to manage relationships with these diverse stakeholders. Some potential strategies include:

• Engage potential competitors in co-opetitive ventures. Satellite operators already depend on terrestrial carriers as distribution partners, suppliers of interconnection to end-users, and customers for wholesale services. Engaging these carriers in such cooperative relationships, where the terrestrial carriers can see large benefits from working with satellite operators, will de-emphasize the tendency of such carriers to undertake destructive competitive practices, such as blocking interconnection or lobbying for trade barriers in broadband telecommunications.
• Develop partnerships with a broad range of technology suppliers. Broadband communications demand will depend on the development of end-user devices and applications which require high bandwidth service, such as ultra-fast data processing, multimedia development tools, and collaborative work applications (e.g., Lotus Notes). Investments or partnerships that accelerate the development and diffusion of such technologies and applications will ultimately boost the demand for broadband satellite services. Intel has used this strategy very successfully -- by investing in software companies whose applications require high-speed processing, Intel has expanded the market for fast microprocessors such as its Pentium product line.
• Work with end-users on issues such as standardization and application development. Satellite operators can form formal or informal consortia of end-users in key application markets, such as education or corporate networking, both to identify key user needs and to develop standards and applications which are compatible with the capabilities of broadband satellite systems.
• Identify key complementors and encourage their development of broadband-based products and services. Broadband satellite operators need to track the efforts of content developers and support their moves to create new broadband content. Such developers include movie studios, television programming producers, interactive game developers, and distance learning providers. Broadband satellite operators can form preferred distribution agreements with those developers, so that the operators' end-users have early access to such content.
• Utilize partnerships with government and university research bodies to support research in multimedia applications and satellite technology. Funding advanced research in multimedia applications will help in the search for the "killer applications" which could spark broad interest in broadband satellite service. One such partnership is already underway in Israel. The Israeli Inter University Computation Center has established a protocol gateway to link its domestic ATM-based OC-3 Internet backbone to a T3 45 Mbps satellite link, operated by Gilat Communications, which in turn links to the international StarTap point of presence in Chicago, Illinois, the connection point to the U.S. Internet2 project (Palter & Nussbacher, 1999). Fundamental research into satellite communications can aid in finding solutions to the technical challenges mentioned throughout this paper. Such research requires both basic study of communications environments and applied research techniques, necessitating a combination of the approaches of government, university, and industrial research (Carayannis and Alexander, 1999).

The market and industry surrounding broadband satellite services will co-evolve with the markets and developers for interactive multimedia applications. Satellite operators will need to look beyond their immediate target customers to identify long-term opportunities to create new markets for broadband service. They will also need to develop innovative approaches to dealing with competitors in broadband communications, in hopes of avoiding price wars and other types of damaging competitive behavior. The first task in pursuing such strategies is to identify significant potential partners outside of the traditional satellite industry, including partners in software, content, computer technology, and Web services, and to initiate relationships with those players that will build grassroots demand for applications that can be delivered via broadband satellite.

References

Brandenburger, Adam M. and Barry J. Nalebuff (1996). Co-opetition. Currency-Doubleday, New York, NY.

Cable, Stephen (May 1998). 'Network bandwidth on demand for multimedia and Internet applications.' Satellite Communications, volume 23, number 5, pp. 48-53.

Carayannis, Elias G. and Jeffrey Alexander (1999). 'An empirical knowledge-based framework for analyzing government-university-industry strategic research partnerships.' Paper presented at 1999 Portland International Conference on Management of Engineering and Technology, July, Portland, Oregon.

Coates, Joseph F. (November/December 1998). 'What picturephone teaches about forecasting.' Research-Technology Management, volume 41, number 6, pp. 7-8.

Coates, Joseph F. (January/February 1999). 'Joe Coates responds...' Research -Technology Management, volume 42, number 1, pp. 13-14.

Dornan, Andrew (July 1999). 'Satellite services: Internet in the sky.' Data Communications, volume 28, number 10, pp. 45-46.

Evans, Andrew L, John S. Rose and Ramesh Venkatamaran (1998). 'The future of satellite communications.' The McKinsey Quarterly, number 2, pp. 6-17.

Hewett, Julian (1997). 'Using scenario-based forecasting to predict future opportunities for new telecom services.' Presented at Market Forecasting in the Telecoms Industry, Institute for International Research, London, December 1-4.

Hudgins-Bonafield, Christy (15 March 1998). 'Networking in the 21st century: the sky's the limit.' Network Computing, volume 9, number 5, pp. 70+.

Hudgins-Bonafield, Christy (15 March 1998). 'The Achilles Heel of next-generation satellites.' Network Computing, volume 9, number 5, pp. 30+.

Lucky, Robert W. (January/February 1999). 'Forecasting information technology.' Research -Technology Management, volume 42, number 1, pp. 12-13.

Mertiens, Markus (1997). 'Determining how to forecast the demand for GMPCS services and the long term potential of global mobile personal communications by satellite.' Presented at Market Forecasting in the Telecoms Industry, Institute for International Research, London, December 1-4.

Mohamed, Ahmed and Frederic Lagacherie (1999). 'IP futures: exploiting the opportunities of a data-driven network evolution.' Presented at ICFC 1999: the International Conference on Forecasting in Communications, Seattle, WA.

O'Rourke, Steve (August 1999). 'Future satellite applications: is the writing on the wall?' Satellite Communications, volume 23, number 8, pp. 22-26.

Palter, PC and Hank Nussbacher (November 1999). 'Internet2 Takes to the Air.' Satellite Communications, volume 23, number 11, pp. 36-39.

Price, David (May 1999). 'Deciphering the hieroglyphics of multimedia by satellite.' Satellite Communications, volume 23, number 5, pp. 34-42.

Stordahl, Kjell (1997). 'Forecasting the demand for broadband services and their influence on broadband network evolution.' Presented at Market Forecasting in the Telecoms Industry, Institute for International Research, London, December 1-4.

Sweitzer, John (July 1999). 'The future of broadband satellite networks.' Satellite Communications, volume 23, number 7, pp. 30-35.