Xem mẫu

The Wireless Data Handbook, ourth Edition. James . DeRose Copyright © 1999 John Wiley & Sons, Inc. ISBNs: 0-471-31651-2 (Hardback); 0-471-22458-8 (Electronic) 6 PUBLIC SATELLITE NETWORKS 6.1 INTRODUCTION Satelliteshavebeencarrying communicationssincetheearly 1960s, serving asrelays forvoice,video,ordata.They havelong been dominatedby the military and, in their direct broadcast form, by large television communications companies. In 1989 the Gartner Group believed that “the mobile satellite industry will be one of the great growth markets of the next decade—surpassing cellular telephones and digital paging, and in the same league as personal computers.”1 Gartner believed that there would be 800,000 satellite data terminals installed by 1993, growing to 2 million by 1995. Afloodof vendors rushedtoward this magnificent opportunity, manyof which were never seen again. After this debacle, intense activity began for new classes of satellites, mostly for voice communications. The successful September 1993 deployment of NASA’s Advanced Communications Technology Satellite (ACTS) showed the potential value of the new designs: on-board processors, spot beams, and Ka band operation to penetrate rain clouds. Hundreds of competitive launches are scheduled for the next few years. By some tallies, at least 1300 satellites are scheduled to be launched into the Ka band alone.2 Iridium was scheduledto beoperational in September 1998. Broadband capability will not be operational until 2002 and includes Craig McCaw’s and Bill Gates’s Teledesic. The enormous broadband capacity is certain to revolutionize some forms of communications. Some market research firms now estimate that just the near-term satellite business could triple to $29 billion per year by 2000. But we have heard predictions like that before. 70 6.2 GEOSTATIONARY SATELLITE SYSTEMS 71 6.2 GEOSTATIONARY SATELLITE SYSTEMS 6.2.1 GEOS Overview Geostationary satellites (GEOS) operate at roughly 22,300 miles altitude, above the equator, where thesatelliteis always over thesamespot on Earth. Becauseof thehuge physical separation between device and satellite, long (1¤2-second) round-trip transmission delays are inescapable. Originally intended for voice, this latency has always been a user problem, requiring speaker discipline. More and more, GEOS are being usedfor data applications where small messagedelaysarenot critical. However, the latency remains a problem for current implementations of some data protocols, particularly TCP/IP. The GEOS ground footprint is huge, capable of communicating with stations in most of a hemisphere. Transmissions generally get through anywhere there is a clear view of the sky in the general direction of the equator but can be blocked in metropolitan areas by buildings. Power requirements are high, leading to physically large and fairly costly ground devices. Geostationary is anexpensiveand difficult orbit toachieve, but successful systems can function with only a single satellite. Only a half dozen sites in the world are capable of launching these large vehicles. About 60% of GEOS launches are controlled by Europe’s Ariane consortium, which is under continuous criticism for high prices and limited capacity. Another 30% of the work falls to McDonnell Douglas and Lockheed Martin, which launch Delta and Atlas rockets from Cape Canaveral and Vandenburg Air :orce Base. The balance of the launches are currently performed by the Russian, Ukranian, and Chinese programs. Struggling to compete are new ventures in Brazil, India, Israel, Italy, and Japan, where Hughes recently signed a $1 billion deal with a Mitsubishi-led consortium for 10 launches. 6.2.2 OmniTRACS In January 1988 a small company named Omninet demonstrated a truck-tracking system that leased time on underutilized GEOS. Because of the size and cost of the device, the solution was only useful for the tractor portion, not trailers. Attracted by this system, Qualcomm absorbed Omninet in August and renamed the offering OmniTRACS. It operated on GSTAR-1’s Ku-band GEOS. With OmniTRACS, Qualcomm began to execute a successful business strategy that had eluded other contenders for 15 years. OmniTRACS slowly rose to the very forefront of the market for AVL-based trucking systems. The device transmission scheme is CDMA, a technology that Qualcomm has since ridden hard and well in the AMPS replacement market. Bit rates are quite low, and facsimile is completely impractical. 72 PUBLIC SATELLITE NETWORKS Studies donein19893 clearlydepict theOmniTRACS goals: schedulingequipment and drivers more profitably, improving driver turnover, and providing better informationto customers. Inspite of aquickstart from Schneider(5000units ordered in 1988 andcompletely installedby August 19894),salesprogress wasinitially slower than expected. At announcement 30,000 installs were anticipated by year-end 1989. This goal was actually achieved three years later. Beginning mid-1993 the pace of installations began to climb rapidly. By the first quarter of 1998 OmniTRACS had shipped (not installed) 230,000 units worldwide, which likely translates to ∼145,000 operational U.S. subscribers. As is foreordained with GEOS, the device is relatively large and expensive (∼$4500). OmniTRACS defends thedevicewith a“Proven HardwareThatRetainsIts Value”5 and “The :irst Mobile Communications Technology Is Still the Best Solution”6 marketing strategy. This seems to be a successful ploy, as worldwide shipments continue their linear growth path. Incremental improvements tothesystemarebeing madeviasoftwarechanges.The area currentlyreceiving the greatestemphasis is theTrailerTRACStrailermonitoring system.7 This approach provides a unique ID for every trailer. The ID is reported to the OmniTRACS tractor unit on every connect/disconnect. This positive trailer ID is a means of tracking trailer assets and provides the ability to monitor trailer pools. Criticalevents are monitored andexceptionreports prepared:for example,lost trailer, excessive trailers at one location, too few trailers, unscheduled movement, wrong trailer connected to tractor, unauthorized cargo, and trailer late for intermodal origin. GPS and time verification of scheduled “drops” and “hooks” are combined with the Dispatch software to track the trailers very closely. But these improvements remain tractor based. In April 1998 Qualcommsigned an agreement8 to deploy Aeris Microburst technology, a terrestrial approach, for the essentially untapped trailer market. 6.2.3 AMSC Skycell :ormed in 1988 from a pool of eight applicants, AMSC began test operations in late 1990by leasingtimeon theGalaxyInMarSat-C system. Trueoperational service was long delayed; AMSC finally got its own SkyCell satellites operational in the summer of 1995. With careful deployment of its spot beams, AMSC blankets all the United States, including Alaska, Hawaii, Puerto Rico, and the Virgin Islands, as well as a great deal of U.S. coastal waters. It is fully interoperable with the Canadian MSAT system. AMSC’s earliest homegrown service, Mobile Messaging, was available in 1993, still via leased satellite time. It offered two-way mobile data and GPS for transportation fleets.9 Customers such as Southeastern :reight Lines (LTL)10 require both voiceand facsimilecapability. AMSC providesthese,aswell as circuit switched data, but separate devices (including antennas) are required for each. There were few early takers of either approach; West Motor :reight11 is an example. 6.3 LOW EARTH-ORBITING SATELLITES 73 The AMSC emphasis has been voice, and it is quite proud of its OmniQuest (how’s thatforname confusion)notebook-sized satellite telephone. Indeed, the quarterlyfinancial reports stress howmuchvoice revenue is generatedper subscriber.Untilrecently,data was not mentioned. Most revenues are still being generated by equipment sales. Service revenues fromthesatellite-only businessmake up only about 40% of thetotal. At year-end 1996 Rockwell “sold” (a cash-free transaction) its multimode fleet management system to AMSC. In this compact AMSC acquired 43 new customers and doubled its data subscriber base with the 8400 units Rockwell shared with ARDIS. This pleasant jolt can clearly be seen in the year-end 1997 total subscriber figures. On March 31, 1998, AMSC completed its purchase of ARDIS from Motorola, stating that it “is positioned to leverage an integrated terrestrial/satellite network.”12 Thefinancial marketsevidently believethis is true.AMSC’s debt offering was sharply oversubscribed and yielded $335 million for the new company. 6.2.4 GEOS Summary Table 6-1 is a summary of the key business characteristics of the two representative GEOS offerings. Notethat eachismoving to cut its absolutedependency onthis form of satellite communication. Qualcomm is moving the trailer application to either LEOS (Globalstar) or terrestrial (Aeris Microburst).AMSC hasa successful terrestrial hybrid system, which is discussed in Chapter 7. 6.3 LOW EARTH-ORBITING SATELLITES 6.3.1 LEOS Overview Low Earth-orbiting satellites operate at 420–1200 nautical miles altitude in polar orbits. Theyarenotgeostationary, and theydo not constantly overlook thesamepoint on Earth. Multiple “birds” must be arranged in an orbital pattern that produces continuous coverage. LEOS are required to have a way of handing off calls from a setting satellite toarising one.Digital techniques are employedthat are wellsuited to data transmission, although the initial focus of most LEOS is decidedly voice. The ground footprint is small to very small, depending upon the system chosen. Spot coverageon the order of 440 (Teledesic) to 2800 (Iridium) miles in diameter is typical. Thelower the orbit,themorethesatellite mustbealmost directly overheadfor signals to get through reliably. The satellites themselves are much smaller and lighter than their GEOS counterparts and can be launched in a number of innovative ways. Orbcomm, for example, saves fuel and avoids the wait for ground launch openings by firing its Pegasus rockets from under the wings of a jumbo jet flying at 40,000 feet. The small satellite size makes possible multiple, simultaneous launchings. Globalstar, for example, plans to launch 12 satellites at a time. 74 PUBLIC SATELLITE NETWORKS Table 6-1 Business characteristics: GEOS networks Service name Parent company Key parent owners Protocol(s) :irst operational Principal emphasis Limitations :irst quarter 1998 coverage U.S. subscribers (AMSC has overlap with ARDIS), year-end 1989 1990 1991 1992 1993 1994 1995 1996 1997 Typical single-user monthly list prices OmniTRACS Qualcomm — CDMA Packet 1988 Trucking/transportation companies: AVL: QASPAR/GPS; nonmetropolitan coverage; long (2000-byte) messages No voice, no facsimile, low (∼150 bps) bit rates Australia, Brazil, Canada, Europe, Japan, Korea, Malaysia, Mexico, USA ∼5,500 ∼8,600 ∼19,500 ∼31,000 ∼54,000 ∼82,700 ∼110,000 ∼133,000 ∼145,000 Monthly access (includes 800 messages), $35; additional messages, $0.15 + $.002/byte SkyCell AMSC AT&T Wireless Polled 1993 Trucking/transportation companies: voice, facsimile, two-way messaging, AVL Polled technique can lead to 4–5 minute latency :ifty States, Caribbean territories, and U.S. maritime waters ∼4,000 ∼20,300 ∼32,400 6.3.2 Orbcomm At least as early as 1989 Orbital Sciences subsidiary Orbital Communications began the design of a LEOS. On August 23, 1993, the 87-pound MicroStar was unveiled.13 The launch date was planned for the first quarter of 1994 with service to begin a few monthslater. The rest of theconstellation wasto bedeployed in four launches of eight satellites each: “By 1995 . . . Orbcomm . . . will be fully operational.” Airtime pricing “will be similar to cellular . . . at wholesale . . . $15 per month.”13 Subscriber devices were expected to “range from $50 to $400, depending on options.” A data system, the primary early adopter was expected to be cargo tracking. The system was optimized for the vehicle location application, with a bright future envisioned for units in cars. ... - tailieumienphi.vn
nguon tai.lieu . vn