Gary L. Garriott
Volunteers in Technical Assistance
1600 Wilson Boulevard, Suite 500
Arlington, VA 22209
In the rush to embrace ever-greater capacities supporting resource-demanding systems such as the World Wide Web, the utility of "real-enough-time" services based on low bandwidth store-and-forward messaging tends to be overlooked. Low-end wireless technologies, including small communications satellites in low earth orbit (LEO) with concomitant low-cost ground-based hardware, can make dramatic impacts in isolated and rural areas, especially in the developing countries of Asia, Africa and Latin America. Such areas are currently left out of the "Internet revolution" as rural teledensity targets (telephones per hundred residents) are not met. These regions may be even further negatively impacted by the trend towards privatization in telecommunications that may view investment in rural infrastructure and connectivity as uneconomic.
Development activities in rural areas are frequently characterized by critical, time-dependent information needs. The time value of information is obvious in rehabilitation and relief response to disasters. Not usually reported, timeliness is also urgent in many standard projects. If real-enough-time access can reduce turnaround query-response cycles from weeks and months to hours or a day or two, this will usually be sufficient.
Volunteers in Technical Assistance (VITA) has been a leader in the use of LEO satellites to provide real-enough-time services. Such satellites in polar orbits pass over all points on earth several times daily, permitting automatic uploading, downloading and delivery when combined with Internet gateway services. A brief history of the first efforts is provided as well as a description of the current attempt to develop high-quality messaging and data services in partnership with Final Analysis Inc., an innovative satellite and network services firm.
Technical descriptions, including operational parameters and use of the Internet as transport for both services, are provided.
The paper concludes that known constraints to widespread adoption of LEO and other wireless technologies enhancing significant impacts in rural development are less technical or human resource-limited than regulatory and bureaucratic in nature.
An answer to a technical problem that takes minutes to obtain in Europe can take months to obtain in Somalia or Sudan. To give just one example, a medical advisor in Mogadishu needed background information on excretion of anti-malarials in breast milk to help him decide on the details of a prophylaxis programme for about half a million people. The agency funding him had no staff in Europe who were themselves qualified to make a thorough search for this information or who knew who to ask to do it for them. The telephone calls necessary to set up and pay for a search through a Western information centre would have taken weeks, given the communication problems at that time. The solution was to get a friend who was passing through Nairobi to pay himself for a search in Europe, personally photocopy the papers concerned, and then to mail the printout and copies of papers to Mogadishu. The total time needed to get the information on this routine enquiry was about four weeks. The programme was already underway when the material arrived. Hundreds of highly technical decisions affecting huge numbers of people are made every month in relief programs with a bare minimum of scientific background data (1).
This is another way of saying that the accuracy of information is an important but insufficient condition for effective use in developing countries. In order for most technical and logistical information to be used in the execution of a project (or in this case, a relief operation) it must be timely as well. Scientists, engineers and physicians--frequently sources of crucial data--are usually aware of the time dimension of technical information requirements in their own professional activities; however, they might not always appreciate that good timing is also highly operative in many relief and development projects alike.
This is not to say that every project always needs information within a few hours or even a few days. Indeed, many do not, but such projects also tend to be based out of urban centers with relatively easy howbeit costly access to the limited national infrastructure. A refugee camp, an agricultural project in a remote valley, a scientist monitoring the spread of AIDS or the movement of locusts through remote sections of Africa all require the ability to transmit and receive technical data and information reliably when the situation demands. These applications require that the communications link be reliably available on short notice without it having to be re-established each time from scratch.
While the paragraph quoted above was written a decade ago, it remains true today, not in the least diminished by progress in digitally-switched-telephone line-mediated Internet connectivity throughout the developing world and indeed in some of urban Africa. One only has to consider what has happened to Somalia in the interim to realize that sole dependency on the terrestrial wire-based communications infrastructure has serious limitations. While especially true for Africa (2), most rural and isolated regions worldwide outside of North America and western Europe are almost completely without connectivity of any kind, including e-mail.
Given the emphasis on expanding bandwidth capacities to accommodate real-time resource gluttons (e.g. World Wide Web), the utility of real-enough-time services based on store-and-forward messaging tends to be overlooked, as has the possibility of providing such services through low-end wireless technologies, including small communications satellites in low-earth orbit with corresponding low-cost ground-based hardware.
Isolated and rural areas of many developing countries in Asia, Africa and Latin America have been and continue to be effectively shut out of the Internet revolution. Two-thirds of people living in the poorer countries have no access to any form of telecommunication and 49 of the world's poorest countries are characterized with teledensities of less than one (0.72) telephone per hundred residents (3). While the International Telecommunications Union predicts average growth in 54 of the poorest countries to 2.17 telephones per hundred residents by the year 2000, the Maitland Commission's 1984 estimate of one telephone per hundred residents by 1988 was dramatically wrong (only 0.72 by 1994). While some of this represents urban back-log that will hopefully be reduced with the advent of newly capitalized digital switch technology, rural areas will still suffer, possibly because of the additional burden of the trend toward privatization. Given sparse populations and dispersed markets in rural areas, there is usually no profit-borne incentive to improve and expand service in these regions by urban-oriented enterprises. Another by-product of privatization, at least in some Latin American and African countries, is that Internet services tend to be provided through the newly privatized companies in a monopolistic fashion. In Belize, for example, even third-party re-sellers of local and international telephone circuits for commercial electronic mail are not permitted as of late 1995 (4).
Such environments do not instill confidence that priority will ever be given to most rural and isolated areas using normative infrastructural standards and policies/priorities. Yet information needs for relief and development are frequently critical, as illustrated in the opening passage above.
Aside from the obvious time-critical nature of disaster and relief situations, information that is crucial to project execution is also time dependent. The same information, if delivered after a certain time, has lost much--if not all--of its value. This is frequently due to the intrinsic value of the information itself, e.g., preparation of proposals and budgets or technical assistance needed at a highly-specific point in time. Even more important is the potential loss of human and material resources that may be used in other projects or wasted altogether if not used when critically needed.
From the standpoint of planners, projects are sometimes viewed as objectives compartmentalized into specific activities, all having discrete beginning and ending points. From the perspectives of field staff, however, it is often more realistic to consider accomplished objectives as having successfully recognized and exploited windows of opportunity. When the window is open, it is critical to have the right information available at that time. When the window is closed, (e.g., field staff have promised skeptical village leaders information on a new treatment for cholera but have not delivered same) it may be twice as difficult if not impossible to reactivate interest. Village leaders, usually characterized with multiple and even extreme demands constantly made on their time, move on to other projects. On the other hand, when information-communication channels are reliable, interesting and unexpected things can happen. A VITA-supported groundstation in Tanzania operating on a LEO satellite (UoSat-3, see below) is currently using VITA's existing Internet gateway to order parts and derive technical assistance for the construction of two small aircraft, built from kits (5). One is already flying and serving health and other needs in remote regions of that country. This use could not have been predicted when the groundstation was first established.
Most useful technical information is the result of multiple pairs of query-response; each response provides more feedback for an ever-refined query. This makes the reduction in turn-around time important and suggests that communication modes that specifically address reliability and speed, particularly from isolated areas, are enormously significant for a great variety of rural projects and activities. Indeed, any conceivable development application will benefit from the availability of this capability. Instant access is usually not required, but if the turn-around is measured in hours (or a couple of days) as opposed to weeks or months, the impact on efficiency and useful function can be dramatic. When one considers that 13,000 international nonprofit organizations plus a plethora of bilateral (USAID and aid-granting foundations from North American, Canadian and European countries) and multilateral agencies (the World Bank, UNDP, WHO, WMO, UNHCR, etc.) are all involved in supporting rural development in some form, a flexible set of technological options permitting "real-enough time" with complementary rural communication policies could have sizeable impacts in these days of diminishing budgets.
Even within communities of experienced Internet users, a curious phenomenon exists. At the very moment when resource demands on bandwidth and computer power are exploding for ever-more sophisticated real-time applications ("Internet telephone" being only a recent example), the capability of accessing much of this same information via e-mail is also increasing. The most comprehensive single source of information on this, "Doctor Bob's Guide to Offline Internet Access," contains nearly 30 pages of resource listings and explanations in the most recent edition (6). Included are references to ftp by e-mail, Archie by e-mail, Gopher by e-mail, Veronica by e-mail, Usenet by e-mail, Usenet searches, WAIS searches, World Wide Web by e-mail, World Wide Web searches by e-mail, e-mail-based mailing lists and "directory assistance." Many other "e-mail extras" exist as well, such as e-mail-to-fax services, dictionary lookups, virus protection software, and even "e-mail-to-snailmail" courier services!
The early advent of LEO satellites for relief and development was championed by VITA (which in 1992 received a "pioneer's preference" toward a nonexperimental permanent license by the Federal Communications Commission [FCC] in the United States). Working with the University of Surrey (Guildford, England), Surrey Satellite Technology, Ltd., the international amateur radio community and its own volunteers, VITA was able to show that communication with LEO satellites (UoSat-2, launched in 1984, and UoSat-3 in 1990) is a technically practical option for developing countries. Modified amateur radio hardware for groundstations was used with specially-developed communications software (based on an X.25-derived protocol), with additional software to control tracking antenna systems operating in the background. Both satellites were launched in polar orbits at altitudes of roughly 800 km, providing for four passes per day at the equator and increasing to fourteen as one approaches the polar latitudes, and at this writing both are still functional. Passes vary in length between roughly 5 and 15 minutes with maximum elevations also changing (but all predicted by the tracking software, which is updated on a monthly basis through the transmission of Keplerian orbital elements from NASA). With UoSat-3 and improved datagram-like protocols, good passes permit the transfer of several hundred thousand bytes of information per pass at the nominal transmission rate of 9600 bps to and from groundstations.
VITA installed or assisted in the installation of more than 25 such ground stations, most in Africa (7), for use on UoSat-3 as well as assisting in the design and financing of the PACSAT Communications Experiment satellite payload. VITA also demonstrated, through a village utility project on several Indonesian islands, that LEO technology could be used reliably in an unattended mode for the transfer of telemetry data (8). SatelLife, a Boston-based organization, has contributed significantly to the experiential knowledge base of using LEO satellites for delivery of health and medical information (9).
This early experimentation and demonstration occurred mostly before the explosion of interest in the Internet registered in the past few years. Installed groundstations communicated directly with other groundstations with transfer of messages destined for other networks via "sneakermail." With the increasingly widespread standardization and availability of IP addressing to public and private e-mail systems alike, VITA devised an innovative means to translate groundstation addresses into their Internet equivalents, e.g., groundstation addressing such as "v.kib" (a VITA-supported station in Kibidula, Tanzania) becomes "email@example.com." A gateway connection between the "vitanet.org" machine and the VITA/Arlington satellite station (physically located within the same office in Virginia, USA) was established by providing the satellite station with a Fidonet "point" address (i.e., 1:109/165.1) which polls the "boss" machine (1:109/165) before and after passes. The packing and exchange of messages occurs using standard Fidonet mailer software. The "boss" machine, also known as "vitanet.org," is equipped with translation software for standard UUCP polling to a local Internet service provider. Sub-domains, such as "sat." in the above address, are assigned by the VITA system operator. Thus, gateway function is fully automatic in both directions, permits unattended operation, uses low cost or public domain Fidonet mailers, is adapted and maintained with in-house expertise, but is not RFC-822 compatible.
In 1992, VITA participated with others (specifically the private companies OrbComm and Starsys interested in the "little LEO" spectrum [frequencies below 1 GHz] for asset tracking and meter reading applications) in the ITU's World Administrative Radio Conference, a regular mega-meeting of this U.N.-affiliated organization that allocates radio spectrum and usage worldwide. The purpose was to persuade the 164 ITU members to allocate slots in very high frequency (VHF) bands and ultra high frequency (UHF) bands for "little LEO" services. The successful result was an allocation for "Non-Voice, Non-Geostationary Mobile Satellite Services," which in turn was created by the FCC and in which VITA and the other "little LEO" applicants were eventually awarded U.S. licenses (after an extensive "negotiated rulemaking" process among the parties imposed by the FCC).
VITA entered into a joint venture with a U.S.-based satellite construction and launch company, CTA, to build and launch a commercial-grade satellite in exchange for access to VITA's license over the United States. This satellite was launched on 15 August 1995 from Vandenberg Air Force Base but unfortunately failed to reach orbit after a malfunction in the Lockheed-Martin rocket and its subsequent destruction by ground crew (10).
VITA now has a new arrangement with another U.S.-based company, Final Analysis, an innovative aerospace engineering and telecommunications firm providing turnkey systems from ground systems and operations, to satellite bus development and payload design and launch services. The satellite containing the VITA communications payload, known both as FAISAT-2v and VITASAT-1R, will be launched aboard a Russian Cosmos rocket in mid-1996, with service initiation near the end of the year. If successful, Final Analysis may eventually launch a constellation of 26 satellites.
From VITA's perspective, the VITASAT communications system contains several elements (see Figure 1). The first is VITASAT-1R (FAISAT-2v) that operates in the "little LEO" VHF/UHF bands with switchable store-and-forward and zero-delay transponder modes. This satellite will be launched into a 1000 kilometer, 83 degree inclination orbit and will provide a minimum of four passes per day for users worldwide (see Table 1 for more satellite specifications). After a second satellite is launched into a similar orbit, Final Analysis plans to launch 24 subsequent satellites in four 67 degree inclination orbital planes of 1000 kilometer altitudes each. All satellites will provide a minimum access of four passes per day per satellite to the VITA system for users virtually anywhere in the world.
The second element is User Terminals (UTs, computers and software) that are attached to Messaging Terminals (MTs) consisting of packet radio modems and radio transmitter/receiver hardware, permitting these stations to send or receive messages from the satellite as short data packets or large files. Messaging Terminals communicate with the satellite at either 2400 bps or 9600 bps (in the United States they are limited to rates of 2400 bps). See Table 2.
In a typical Messaging Service scenario, a UT is connected to an MT via a local area network over Ethernet or "terrestrially" via packet radio protocols. The MT receives a message from one or more UTs and sends this message to the satellite. The satellite receives the message and routes it to a satellite gateway (currently, planned for Andenes, Norway, and Capetown, South Africa) where routing decisions are made to deliver it to a local UT or to send it through the corresponding Internet gateway. If the destination is a remote UT, then a return path back through the satellite gateway to the satellite and to the proper MT will be made.
In addition to MTs, Remote Data Terminals (RTs) will be used for the transfer of remote telemetry and sensor information for meter reading and asset tracking. RTs are self-contained units (no computers attached, unlike the MTs) that transmit and receive short, variable-length packets via User Ground Stations (UGSs), which are a functional subset of the Master Ground Station (MGS) and Network Control Center (NCC) described below. When UGSs and RTs are located within the same footprint, they can use the zero-delay transponder mode. This mode precludes the need for store-and-forward re-transmissions of signals outside the footprint. UGSs communicate with the Data Service customers using leased lines, public switched networks (e.g., X.25, frame relay) or the Internet.
The Master Ground Station (MGS) controls the operation of the satellite constellation, while the Network Control Center (NCC) handles the data flow and controls all Internet gateways. In addition, the NCC performs the billing and customer service function, including technical support. It will be located at Final Analysis offices in Beltsville, Maryland. Satellite control is performed at the MGS located at Final Analysis facilities in Logan, Utah. In an emergency, functions of the MGS and NCC can be interchanged.
As the satellite orbits the earth, it scans the uplink band (149.810-149.9 MHz) at the start of every TDM (time-division multiplexed) frame. (Over the United States, frame activity begins sweeping in 2.5 kHz steps throughout the uplink band). The frequency of a clear channel is then identified and broadcast to the MT as the uplink frequency to be used. Those MTs with messages to uplink then request time slots within the TDM frame. When the request is acknowledged by the satellite, the MT uplinks the message in packets of up to 512 bytes in length (including overhead). This arrangement permits MT-initiated requests up to a maximum of 7 percent of uplink capacity. All other uplinks are RT uplinks, commanded by the satellite through the downlink, since each satellite in the constellation will maintain location and addressing information for each MT and RT over which it passes.
E-mail packets are stored in the satellite's mass storage processor until the satellite passes over a Satellite/Internet gateway station. As the satellite appears over the horizon, the station will request and communicate over a dedicated downlink channel, while MT and RT traffic to the satellite within the footprint of the satellite continues. The station will then command the mass storage processor to downlink the messages it has collected at either 19.2 or 38.4 kbps data rate. Next, it will uplink at 19.2 kpbs any messages to remote users that have arrived at the gateway in time for that satellite pass. At the end of this sequence, the station will release control of the dedicated downlink channel, allowing any other fixed stations (such as UGSs) within the footprint to access the satellite through this channel. Routing to the Internet from the gateway then occurs as previously described (see Figure 2).
The Data Service shown in Figure 3 has been designed as a cost-effective method for commercial and industrial customers in the United States and elsewhere to gather data from sparsely populated areas. This service allows users to send small packets of data from RTs used in applications such as environmental monitoring, asset tracking, and utility data collection. The Data Service can use UGSs located in the same footprint as the population of RTs it will serve. In addition, the Data Service can use the Satellite/Internet gateways to downlink data collected from areas where there are no UGSs. While any given UGS will receive a large volume of data, it will transmit relatively little. All UGSs and RTs in the United States operate under the same restrictions as the MTs (i.e., packet transmission spacing of 15 seconds at any given frequency and a burst length maximum of 450 msec as well as a maximum of 20 transmissions over a 15-minute interval, all at 2400 bps). Like the Internet gateways, these hubs are controlled by the NCC. Unlike them, however, the UGSs have direct links to the commercial and industrial customers for whom the data are being collected.
The RTs can collect data from a single source or may act as concentrators for data collected from multiple sources through terrestrial packet radio or power line transmission systems. In either case, as the satellite comes into view of an RT population, the satellite begins the data collection sequence by performing a bandscanning and frequency assignment sequence in the same manner as the Messaging Service. Clear channel frequencies will be assigned to the RTs, which will then adjust themselves accordingly. At this point, the satellite will transmit interrogation commands to the RTs. Data from RTs will then be received by the satellite for relay to a UGS, the NCC, or an Internet gateway.
All RTs and MTs will be given unique addresses under a proprietary scheme permitting the existence of about four billion separate units; these addresses are translatable into their IP equivalents. Each unit will also have a proprietary group address, so that sites in a common geographical area can be addressed together. TCP/IP (UDP/IP) packets are encapsulated within proprietary "Final Analysis packets" for node-to-node transmission. The VITASAT-1R satellite itself has been assigned a Class C IP host address (126.96.36.199 or faisat.facs.com). The entire system will use all standard TCP/IP schemes for addressing, routing, and multiplexing different hosts and networks. In the LEO world, this flexibility can be important for special applications. For example, small African radio stations could benefit from IP broadcasting for delivery of "newswire" information and/or newsgroups to which they otherwise would not have access, as the cost of standard media news services is prohibitive. MT design will also be integrated with standard IP-compatible "ground" protocols, e.g., ethernet, NOS/KA9Q, etc., for maximum interchangeability.
From modest beginnings existing prior to the Internet revolution, VITA has persisted in the quest to bring unserved areas into the promise of the information revolution through low-cost, radio-based technologies. Initially overlooked as primitive, low earth orbiting satellites when combined with the flexibility offered by TCP/IP protocols over the Internet today offer a rich suite of modern technological solutions for a wide variety of information requirements. While essentially store-and-forward, the increasing spectrum of e-mail-based services available in "real-enough time" greatly magnify the utility of such low bandwidth systems.
Constraints to deployment of this technology are not now believed to be technical in nature. Costs should be reasonable, perhaps as low as $1,000 for mass-produced MTs and USD 50 for 100K of monthly message transfers, with lower costs in bulk transfers exceeding this amount. Given many years' experience in training technicians and engineers from developing countries in radio-based digital technologies (11), innate abilities to handle advanced technical material (especially with regard to entrepreneurial situations) exist and are not viewed by VITA as constraints either.
The real concern is how nations will license and regulate LEO satellite operations and wireless technologies in general, since traditional monitoring required when security concerns are paramount is mostly rendered meaningless. Liberalizing the regulatory apparatus will probably require a national willingness to analyze distinct political, economic, legal and cultural milieus with regard to specific needs addressable by
information-communication technologies. A "cookie-cutter" approach to this process is generally not possible and must be developed on a country-by-country basis, and, even better, region-by-region.
VITA continues to hope that the international development community and national governments will recognize the benefits from net human capital inflows of knowledge and expertise enhanced by such communication systems when encouraged without excessive bureaucratic interference. The ITU will convene the First World Telecommunication Policy Forum in Geneva in October 1996, which will set forth global policy principles for global satellite systems, including the "little LEOs." LEO messaging systems, as one element in a panoply of wireless communications technologies, can provide the long-hoped-for technological
"leap-frogging" toward bringing rural areas into the mainstream of development as the world hurtles toward the 21st century.
The author wishes to acknowledge the substantial contributions made by Louis Ruffino, Tony Sanders, Rob Atkin, and Mary Kay Williams, all of Final Analysis, in the preparation of this paper. A special thanks goes to VITA staff Barbra Bucci who prepared the figures on short notice and to Henry Norman for his extensive review and comments.
Size: Basic structure is 90 cm x 40 cm x 1.1 m high
Weight: 90 kg [198 lbs.]
Gravity gradient boom for attitude stabilization
On-board GPS receiver for attitude determinations
Deployable folding solar panels
Loral RS 6000 (RISC) Command and Data Handling (C&DH) processor
RS 6000 multi-tasked with C&DH processing tasks
16 mbytes of error-corrected message storage memory
Operates at data rates from 2,400 bps to 38,400 bps
2 active receivers, scanning overlapping channels in the 148.00 - 149.900 MHz band
1 active transmitter plus one spare, both capable of 15 watts output tunable over the entire 400.0 - 401.0 MHz band
nominal downlink frequency: 400.55 MHz
Right-hand circularly polarized antennas for uplink and downlink
Altitude: 1000 km
Orbital Period: 105 minutes (time to make one complete pass)
Average single pass viewing time (with 10 user horizon): 10 min.
Footprint of satellite (With 10 horizon): 4800 km
Minimum number of passes at the equator in 24 hours: 4
Minimum amount of visibility at the equator per 24 hours: 40
Pass Spacing: 26 degrees west per pass
Launch Vehicle: Cosmos 3M
Launch Site: Plesetsk, Russia
Launch Date: Second Quarter 1996
Satellite licensed by: Federal Communications Commission (USA)
ITU Radiocommunication Bureau (IFRB/RB) filing: Leotelcom-3, RES46.C.43 (also referred to as FAISAT-2v)
Frequency Band: 149.810-149 MHz
Tunable Step: 2.5 kHz
Bandwidth: 25 kHz
Transmitting Power: 10 W
Modulation: GMSK/9.6 kbps
Transmit Duration: max 450 msec
Duty Cycle: <3 seconds per 15 mins (in U.S.)
Frequency Band: 400.15-401 MHz
Bandwidth: 30 kHz
Demodulation: GMSK/4.8 kbps
Noise Figure: 3 db
Data Acquisition: RS-232
Memory: 4 MB
Modes: Off, Sleep, User Access, Standby, Full
RS-232 Data Port
External Power Source