Towards a policy framework for integrating space technology applications for sustainable development on the information superhighway

3. Space technology on the information superhighway:
broadband access through satellites

Space technology applications today extend over meteorology, communications, broadcasting, resource management, navigation, health, environmental management and disaster management, thus touching virtually every facet of human endeavour. Space activities are expected to be one of the major areas of growth in the twenty-first century, driving technological developments in numerous fields ranging from telecommunications, telematics, multimedia, opto-electronics and robotics to life sciences, energy and nanotechnology. Space business will also see enormous investment in space systems and ground infrastructure, and a rapid increase in downstream market applications. Space applications have proved their indispensability in many applications of remote sensing, communication and navigation, and in the coming days they will continue to shape our daily lives and enrich our knowledge about the universe and our own Earth.

As the convergence of communications and information technology sweeps the world, and as access to the Internet is considered an essential first step towards bridging the digital divide, it is important that we briefly review the space system scenario in terms of its role in providing global connectivity and its contribution to “global well-being”.

3.1 Space technology for digital bridging

Even as the development of ICT applications promises a ray of hope to the developing countries, the prospect of accelerated growth still seems to be far away, because less than 5 per cent of the global population is engaged in activities related to ICT. The digital divide continues to pose a major challenge to many of the developing countries, which are still grappling with a severe shortage of telephone lines, lack of electricity and high levels of illiteracy. Almost 2 billion of the world’s people have no access to energy utility services, and even though at least one billion of these are able and ready to pay for these services, we are not able to reach them. Clearly satellite technology has a great potential to reach these people, and in some cases, it is the only technology that can provide connectivity to remote or difficult-to-access regions (Othman and Kotelnikov, 2001). Satellite systems play an important role in enhancing the ICT landscape, extending necessary services to the hard-to-reach and bridging the digital divide.

There is no question that satellites, from their vantage points in low, medium or geo orbit, provide a synoptic view and coverage of the vast land mass, whether it is for resources management or for global connectivity. Space technology, through communication and remote sensing satellites, contributes to both the conduit and the contents for the evolving information superhighway. While remote sensing and Global Positioning System (GPS) satellites provide data on natural resources and the environment from which we may generate information (and hence increase the value of the resources), the communication satellites provide necessary interconnection to information sources, thus enabling access to “information services” themselves.

Integrating space technology applications into the information superhighway demands an understanding of the technology and the underlying issues, even as we try to formulate an appropriate policy framework in tune with the fast-changing technologies and the growing, new demands and expectations from the users. Chapter 2 lists various enabling technologies, both wireline and wireless, that can be used as vehicles to reach the users. This chapter reviews the status and future scenario pertaining to the broadband satellite segment in particular, which is set to provide affordable Internet access to users across the globe in the coming years.

3.2 Broadband access through satellites: advantages and challenges

Recent years have seen an exponential increase in the demand for high-speed networking. The expansion of Internet services and the rising expectations of the user are the primary reasons for this demand for broadband connectivity. Technology has also made it possible, thanks to the increasing capacity of semiconductor chips available at ever-falling prices. As a matter of fact, the very notion of what is broadband has changed over time. While 64 Kbps was considered broadband a few years back for VSAT networks, it is no longer the case today. Newer applications are consuming available bandwidth and are further driving the need for still higher rates. Currently, broadband connotes rates of 1 megabit per second or higher, and it is likely that this definition will change with time as more and more capacity is added, in both terrestrial and satellite segments. The next generation of satellite systems will have a total capacity in the gigabit-per-second range, while future systems are claiming capacities in the terabit range. Clearly, satellite systems cannot compete with the capacities achieved in terrestrial fibre optical systems. However, the unique features and advantages of satellite systems in the overall context of access and contents guarantee an increasingly important role in many applications (Hadjitheodosiou and others, 1999).

Much of the broadband connectivity is provided currently through terrestrial networks. But with increasing demand for services such as multimedia videoconferencing and software distribution across many sites that are widely dispersed, satellites are well suited to carry such broadband applications. Their ability to serve many users and to solve the expensive “last mile” issue (making it unnecessary to dedicate cable, fibre, switching equipment ports etc. to every user) makes satellites attractive for broadband communication. Similarly, interconnecting geographically distributed high-speed networks through satellites is yet another attractive proposition.

The increasing popularity of broadband satellite systems stems from two major trends in the market. First, terrestrial broadband build-up, or spread, is a slow and costly process, leaving a significant number of users uncovered. Broadband satellites are expected to fill the gaps in the terrestrial broadband infrastructure. Second, congestion in the terrestrial network does not allow the free flow of more complex Internet content such as streaming broadband file distribution and complex multimedia. Satellites bypass the terrestrial congestion by distributing the content from source to edge location via one hop, avoiding multiple routers and terrestrial exchange points.

Thus, rather than being a replacement for the terrestrial telecom networks, satellite networks actually complement the fibre optic backbones, bypassing congestion points and eliminating the “last mile” problems. Satellites can deliver information directly to an office local area network (LAN) or caches and servers placed at strategic points at the edge of the public broadcast networks. Today, xDSL, cable modems and microwave distribution networks simply cannot reach every potential customer, even in the most developed countries, let alone the less developed countries in Asia and the Pacific, where adequate infrastructure does not exist. Added to this, the great cultural and linguistic diversity of the region makes the need for multimedia services more acute. Yet another problem with most public broadcast services lies in the nature of today’s Internet Protocol-based services. Until there is much more bandwidth available globally, it will be difficult and also expensive to provide multimedia applications services over the public IP backbone and maintain the high-quality services expected of multimedia applications. Another difficulty with terrestrial broadband networks is that the network environment changes from one service provider to the other. Different tariffs, proxy servers, dynamic IP addresses, and different qualities of service make it a nightmare to implement web applications over multiple sites in multiple countries. On the other hand, broadband satellite networks provide a single-network architecture to reach every remote site with consistent quality of service, regardless of local conditions (Gilroy, 2002).

Essentially, the broadband communication satellites are recognized as the platform of choice for certain Internet Protocol applications, calling for high data rate connections. Multicasting, global coverage, and ubiquity of service are the core advantages of satellites in the access and content market. The inherent benefits of these satellites include the following (Northern Sky Research, 2001):

(a) One-to-many support: Potentially millions of users are served through the same downlink, thus achieving efficiency and quality of service considered impossible by terrestrial distribution methods;
(b) Ubiquitous, global coverage: A network of three or four GEO satellites offers virtually global coverage, an important advantage for regional access services and IP content distribution applications to multiple, dispersed locations;
(c) Rapid installation and service provisioning: A two-way satellite dish and a gateway can be installed in less than half a day. Mobility and redeploy capability are the associated advantages;
(d) Asymmetric bandwidth support: The uplink and downlink bandwidths are configured at different data rates. This benefit mirrors the asymmetric nature of most Internet traffic – a narrowband outbound Internet request is satisfied by a broadband downlink to the end-user;
(e) Bandwidth-on-demand: Satellites are also uniquely designed to support emerging bandwidth-on-demand applications, whereby the end-user pays only for the bandwidth consumed.

Broadband connectivity through satellites is also attractive because of the fact that network expansion is not a significant planning exercise, as new users can be accommodated simply by installing new Earth stations at the customers’ premises. Satellites also support a wide range of customer bit rates and circuit provision modes, besides enabling provision for common alternative channels for routes where demand and traffic characteristics are uncertain, thus maximizing the resource use efficiency. With these advantages, there is no doubt that the broadband satellites will play an important part in the evolving global information infrastructure.

3.3 Spectrum considerations

As discussed above, the increasing need to accommodate high-rate transmission in satellite systems pushes the usable spectrum into higher frequency bands, such as Ka-band (27-40 GHz) and V-band (40-75 GHz), since the Ku-band (12-18 GHz) and C-band (4-6 GHz) are unable to accommodate such high data rates. Most VSAT and DBS television systems in operation today use portions of the Ku-band. There is simply not enough spectrum to accommodate all the terrestrial and satellite-based wireless systems in a single band. The main problem with the Ka-band is that the signal is lost in rainy conditions. Continuing demand for additional bandwidth has forced commercial satellite system operators to consider the V-band as well; some military systems already operate in this frequency range. These higher frequencies offer additional challenges due to severe multi-path fading and scattering of transmitted signals (Hadjitheodosiou and others, 1999).

3.4 Broadband satellite applications

With broadband satellites serving as the conduit in the information superhighway, a diverse number of applications could be served better. Some of the applications of relevance to the region and identified as part of the Minimum Common Programme of the Regional Space Applications Prgramme for Sustainable Development (RESAP) which could be serviced by broadband satellites are distance education, telemedicine and disaster management applications, not to mention a host of other applications pertinent to decision-making. For example, satellite data relay services, videoconferencing, streaming audio and video, video telephony, digital audio broadcasting, transmission of business transactions, and connection of private business intranets are among the main services proposed by the next generation of satellites. WorldSpace has already started digital audio broadcasting services.

The “killer application” of the broadband satellites is essentially high-speed Internet access. Among many possible services, the services best suited for broadband satellite systems and that are relevant to the current study are (a) asymmetric TCP/IP to support Internet applications, (b) multicasting applications, and (c) web page caching. While the provision of access to the Internet in future broadband satellite systems will not be exclusively asymmetric, for reasons more economic than technical, such access is likely to be asymmetric for most individual users. With regard to multicasting of information to a large number of dispersed users, there is likely to be a large demand for this type of service from broadband satellite systems in the future. Satellite-based videoconferencing could be established by tunnelling IP multicast messages through satellite gateways, but this would require establishing multiple tunnelled virtual circuits between geographically separated users. Efficient use of satellite constellations for such group applications also calls for satellite onboard switches that include support for multicasting (Hadjitheodosiou and others, 1999). Web caching relies on storing the data in caches at the “edge of the Internet”, which speeds up delivery and relieves congestion points along the network backbone. (See also section 3.6 in this chapter, “Connecting to the Internet edge”).

3.5 Broadband satellites: market scenario

In spite of the fact that a single fibre-optic cable can carry more than 640 Gbps, which is more than the total throughput of all GEO satellites currently in orbit (estimated to be 250 Gbps), more than half of the countries relying on IP services will look towards broadband satellite services. It is forecast that the global market for broadband satellite access services will increase from US$ 330 million in 2001 to more than US$ 12 billion in 2006, while during the same period the satellite multicasting and content delivery market is forecast to grow from US$ 160 million to US$ 3 billion (Northern Sky Research, 2001). Satellites are already carrying some 15-30 per cent of the international IP backbone, which is estimated to be around 30 GB. Around 235 satellite transponders are providing Internet-via-satellite, and this is expected to grow to more than 875 in 2004. Market studies indicate that the transponders’ lease charges for the Internet services are expected to be around US$ 710 million in 2001, three times larger than in 1999. In the Asian and Pacific region, the total satellite bandwidth is expected to grow from the about 5.4 Gbps in 2000 to more than 21 Gbps by the end of 2003. In terms of revenue, Intelsat, the market leader in satellite-based backbone IP traffic, generated US$ 90 million in IP-related revenue and managed more than 3.4 Gbps in IP traffic, including 800 Mbps in the Asian and Pacific region alone. Intelsat quadrupled its business in one year, between 1998 and 1999 (Othman and Kotelnikov, 2001).

Recent market surveys (Northern Sky Research, 2001) indicate the following:

(a) The primary opportunities for broadband satellite services come primarily from broadband access and content distribution and delivery services. Because of strong competition from terrestrial systems, satellites fill in where terrestrial networks are lacking. The market opportunities will be in a diversified series of niches centring on the primary applications (access and multicast);
(b) Ultimately, the broadband satellite market will depend on the quality of service at the right price point and on technically sound equipment;
(c) Direct Broadcast Satellite is the key to residential broadband subscribers. Offering bundled broadcast and television services through a single antenna seems to be the future trend;
(d) The satellite equipment market appears to be split between the digital video broadcasting-return channel via satellite (DVB-RCS) proponents versus proprietary solutions vendors with mainly TDMA or CDMA or the demand-assigned multiple access (DAMA) return channels. DVB-RCS proponents claim that this standard will help the industry through lower costs, customer choice and equipment interoperability. However, the proprietary vendors, who already have a dominant chunk of VSAT and the one-way residential market, have little interest in the open standards;
(e) For pure two-way connectivity play, Ka-band appears to be the best option for satellite access platforms, but this trend hinges on solving the rain fade issue associated with Ka-band. In the meantime, the multicast services that require little or no return channel may continue with Ku-band in the short to medium term;
(f) The “bent pipe” versus onboard processing debate continues. When a satellite using the bent-pipe system receives a signal, the signal is retransmitted without any processing. While the bent-pipe satellites have performed quite well in the first phase of Internet development via satellite, the onboard processing design option is being tried by the proposed Astrolink and SpaceWay. Better bandwidth efficiency and true mesh connectivity may tip the scales in favour of this technological advance;
(g) Hybridization of networks is perceived to be the best option to optimize efficiency and improve the overall level of service;
(h) Despite the latency issues, GEOs appear to be the most economically feasible option for delivering broadband satellite services, having been proven for reliability. Most LEO configurations require high up-front costs and an equally complex terrestrial network for routing and interconnection. Over 40 orders were placed for GEO satellites in 2000, which is at least a 30 per cent increase from 1999.

The LEO systems, such as Iridium, Globalstar, ICO, SkyBridge and Teledesic, could not really take off because of financial difficulties. Of the above, the first three are essentially meant for voice with limited Internet access, albeit at rather low bit rates. Iridium went into bankruptcy when its projected market volume failed to materialize. Currently, it is continuing its limited operation only at the behest of the United States military. Globalstar is facing a similar situation. ICO, SkyBridge and Teledesic have also had problems receiving enough investment to proceed with their original schedule. Teledesic also scaled back the number of satellites from 840 to 288.

Satellites in GEO, as seen earlier, seem to be faring better, and an increasing percentage of their capacity is being used to carry Internet traffic. This capacity is primarily used to connect Internet service providers (ISPs) to the Internet backbone, rather than connecting individual users to an ISP. Therefore, in many parts of the developing world where there is a scarcity of residential phone connections, these satellites cannot provide the type of access that is really needed. There are new, very small aperture terminals, with apertures of less than 1 metre, appearing in the market from Gilat/Staband and Hughes Network Service in the United States, which can provide direct access to individual users. They still cost in the US$ 500-1,000 range, not affordable for many individual consumers. The broadband GEO systems such as Astrolink, SpaceWay, EuroSkyWay and others have also announced plans to reduce the number of satellites that will be initially developed. There may be very little capacity available from these satellites, at least initially, to developing countries (Ashford, 2001).

L-band satellites such as those of Inmarsat and the more recently launched regional satellites ACeS and Thuraya cover a large part of the Asian and Pacific region with the capability of connecting individual users to an ISP, but generally only at relatively low data rates and high prices. For example, at present, Inmarsat operates a Global Area Network (GAN) providing a mobile ISDN service at 64 Kbps. It also offers 144-Kbps regional services to laptop-sized terminals through the Thuraya satellite, which covers an area from West Africa to Central Asia. The latter is by way of preparation for Inmarsat’s own broadband GAN (B-GAN), expected to go on stream by 2004 using the fourth generation I-4 satellites. It will offer services at up to 432 Kbps. Inmarsat’s “building block” approach seems to be one of the more practical, because it builds on a reliable global network by adding new satellites with higher capacity and greater flexibility as the market develops (Williamson, 2001).

The trend in the cost per megabyte via satellite is, however, one of slow yet continuous reduction. While the cost of operating the large LEO constellations limits their degree of further cost reductions, regional GEO systems should be able to reduce their costs significantly. For the future broadband systems, prices in the order of US$ 0.10 per megabyte have been predicted. Ultimately, the satellites will prove to be the best near-term solution to providing global interconnection to information sources, but only if some simple conditions are met (Ashford, 2001):

(a) A suitable satellite system is available, and is inexpensive enough to be affordable by the developing countries;
(b) People have means to connect to the Internet via satellite;
(c) The information is provided in the local language.

While satellite broadband is a solution to the vexing question of two-way Internet connectivity, the current cost per megabyte is still high, and satellite solutions are used in places where alternatives are not available. Currently, the high terminal and installation charges in addition to the limited bandwidth capabilities make the service viable only to a small segment of the user community. From a cost perspective, fibre-optic systems deliver 1,000 times the bandwidth of broadband satellite systems for about the same deployment cost. It is, therefore, natural that fibre will continue to dominate in providing the Internet backbone and transoceanic broadband capacity. However, satellites will play their true role in larger geographical areas where there is a need for broadband network access and where there is no terrestrial infrastructure available. The basic technology categories that form the basis for various satellite broadband offerings could be grouped as Ku-band fixed satellite service (FSS), bent-pipe Ka-band, onboard processing Ka-band, and L-band mobile satellite services (MSS). From a practical standpoint, Ka spot beams provide 30 to 60 times the system capacity of the FSS approach, and this helps to make satellite broadband services a long-term, economically viable business opportunity, as end-users‘ bandwidth requirements will only increase over the next 5-10 years (Puetz, 2001b).

Table 3.1 Representative broadband satellite systems

For these reasons, broadband satellites are expected to consolidate their position further, once Ka-band services are operationally available. With the high-speed Ka-band satellites – WildBlue expected in late 2002, and Hughes’s SpaceWay and Shin Satellite’s iPStar in 2003 – promising competitive prices, the market scenario will undergo vast changes in the next two to three years, posing tough competition for the terrestrial cable modem and DSL operators, particularly in the rural areas, where the infrastructure is weak and fibre networks are not likely to penetrate for a long time to come. With transponder numbers increasing, the overall added bandwidth is slated to go further.

Figure 3.1 depicts a scenario of satellite bandwidth as a percentage of Internet bandwidth in the coming years.

3.6 Connecting to the Internet edge: leverage of satellite-based delivery networks

Delivering multimedia content over the Internet has brought about challenges that were not foreseen by the content providers who used to host their sites on a single web server and served the entire globe with low-bandwidth content. However, as the content became richer, demanding larger bandwidth, many servers and the network were not able to support the increased demand of the broadband content, resulting in congestion of the costly international backbone. Basically, there are two approaches to solving this Internet performance issue: (a) increasing overall capacity, performance and reliability by adding adequate infrastructure, and (b) providing an overlay network that bypasses the congestion points. The second approach is the one used in the content delivery network in which the contents are moved as close to the user as possible, so as to reduce the access latency and increase service reliability and throughput. In simplest terms, this approach replicates the content and serves it from the “edge” of the Internet and is often less expensive than serving the content from a small number of central locations, because bandwidth costs reduce at the edges, and the cost of data storage is much less expensive than communications bandwidth. In general, the content delivery networks (CDN) use a combination of distributed caching, load balancing and web request redirection systems. Based on user proximity and server load, content is served from a cache located at the edge of the network, a single router hop from the user (Puetz, 2001a; Yen, 2001). CDN architecture can be implemented either through the terrestrial network or through a satellite. The complexity and the cost of terrestrial CDN increases proportionally with the number of edge servers. For a region like Asia and the Pacific that has a very poor cross-connect infrastructure between countries and where most of the ISPs have transit connectivity in the United States, this means that many router hops are needed to connect the CDN hub to the edge server, creating inconsistent latency, uncontrollable packet loss and difficulty in managing the network. On the other hand, a satellite-based CDN completely bypasses the terrestrial congestion points and provides connectivity to any number of edge servers without increasing the cost, thanks to its point-to-multipoint nature of transmission. Among the various approaches, the use of a hybrid terrestrial-satellite approach has been preferred because of the huge economic and throughput benefits of multicasting via satellite.

Taking advantage of this approach, many enterprises that are looking for cost-effective, technically sound IP broadband solutions are opting for satellite networks for accessing the remote offices. The long-term projection for the enterprise broadband access and VSAT networking, as well as IP multicasting and content delivery, indicate a positive trend in the coming years, as table 3.2 shows.

Table 3.2 Enterprise IP via satellite services market (billions of US dollars)

Source: Northern Sky Research, 2001. Broadband Satellite Markets: A Comprehensive Analysis of Trends and Opportunities. <www.northernskyresearch.com>.

Similarly, the consumer broadband satellite revenue projections also show a positive trend in the coming years (table 3.3), particularly from 2004 onwards when the Ka-band systems would have been operational.

Table 3.3 Consumer broadband satellite revenue (billions of US dollars)

Source: Northern Sky Research, 2001. Broadband Satellite Markets: A Comprehensive Analysis of Trends and Opportunities. <www.northernskyresearch.com>.

3.7 Regulatory issues

Besides the spectrum-related issues calling for extensive coordination with the International Telecommunications Union, which is the international authority for the management of the electromagnetic spectrum at the national and regional levels, a company developing a satellite system must secure permission not only from the home government to service stations in other countries but also from those countries as well. There could be concern in those countries with regard to potential interference with terrestrial systems or the content to be delivered on the proposed satellite system, on the grounds of the content being politically, morally, culturally or religiously offensive. Some of these concerns have strong implications for the technical character of the satellite system itself, such as modifying satellite antenna radiation patterns. There could also be proprietary concern with regard to the security of the content itself.

Countries do realize that satellites have a major role to play in the world’s communication infrastructure. The successful and harmonious integration of satellite and terrestrial systems poses challenges, both in technology and at the policy level, that the national Governments need to overcome. Making an appropriate technology choice is itself a major task in most of the countries. It should be recognized that, even as the technology progression has enabled fusion of various technologies on the information superhighway, and as the globalization of trade has opened new vistas, drawing regulatory boundaries to deal with the rapid expansion of non-traditional players, who could be from across the world in the globalized environment, is not an enviable task for most countries, and even less so for the developing countries. With multiple players operating at different levels across the countries in the value chain, such as the content creator, service provider, satellite operator and the access provider, there will no doubt be many grey areas in effecting control of or making regulations concerning content, access or any other issue. For these reasons, the setting up of an effective regulatory mechanism is a complex issue. Many a time, a code of conduct or a set of guidelines to be followed in the concerned countries urging the players to be sensitive to the values and cultures prevailing in the country or region has worked reasonably well. For example, in the area of satellite broadcasting it has worked, with the players adopting a responsive approach in determining acceptable and unacceptable content (Sadhu, 2001).

The Asian and Pacific region has the highest number of regional satellite players in the world, with virtually every country backing at least one national satellite operator, who usually enjoys preferential treatment within the host country, thus keeping many service prices high and competition low. Added to this, the Asian and Pacific region has many regulatory barriers such as restrictions on importing satellite technology, blanket licensing, terminal approvals, connections to the public telephone network and service licenses, which limit the growth potential of Internet via satellite services (Baugh, 2001). There are also issues related to coordination among various agencies within the country itself. There are other issues related to the life cycle, costs, standards, modularity and interconnectivity, interoperability, technology transfer and human resources development, all of which need appropriate attention while the policy framework is being constructed. Market liberalization has definitely progressed well in the past few years in many countries in the region. If they are left behind with an inappropriate policy framework in this convergence race, some countries run the risk of being sidelined by the global economy itself, besides denying their population the chance to “leapfrog” in technology and development. International agencies such as ESCAP have a major role to play in bringing this awareness to the developing countries and enabling them to acquire appropriate state-of-the-art technologies at affordable costs.