Kayton, M. “Navigation Systems” The Electrical Engineering Handbook Ed. Richard C. Dorf Boca Raton: CRC Press LLC, 2000 109 Navigation Systems 109.1 Introduction 109.2 Coordinate Frames 109.3 Categories of Navigation 109.4 Dead Reckoning 109.5 Radio Navigation 109.6 Celestial Navigation 109.7 Map Matching Navigation 109.8 Navigation Software 109.9 Design Trade-Offs 109.1 Introduction Navigation is the determination of the position and velocity of a moving vehicle on land, at sea, in the air, or in space. The three components of position and the three components of velocity make up a six-component state vector that fully describes the translational motion of the vehicle because the differential equations of motion are of second order. Surveyors are beginning to use the same sensors as navigators but are achieving higher accuracy as a result of longer periods of observation, a fixed location, and more complex, non-real-time data reduction. In the usual navigation system, the state vector is derived on-board, displayed to the crew, recorded on- board, or transmitted to the ground. Navigation information is usually sent to other on-board subsystems; for example, to the waypoint steering, engine control, communication control, and weapon-control computers. Some navigation systems, called position-location systems, measure a vehicle’s state vector using sensors on the ground or in another vehicle (Section 109.5). The external sensors usually track passive radar returns or a transponder. Position-location systems usually supply information to a dispatch or control center. Traditionally, ship navigation included the art of pilotage—entering and leaving port, making use of wind and tides, and knowing the coasts and sea conditions. However, in modern usage, navigation is confined to the measurement of the state vector. The handling of the vehicle is called conning for ships, flight control for aircraft, and attitude control for spacecraft. The term guidance has two meanings, both of which are different than navigation: 1.Steering toward a destination of known position from the vehicle’s present position, as measured by a navigation system. The steering equations on a planet are derived from a plane triangle for nearby destinations and from a spherical triangle for distant destinations. 2.Steering toward a destination without calculating the state vector explicitly. A guided vehicle homes on radio, infrared, or visual emissions. Guidance toward a moving target is usually of interest to military tactical missiles in which a steering algorithm assures impact within the maneuver and fuel constraints of the interceptor. Guidance toward a fixed target involves beam riding, as in the Instrument Landing System, Section 109.5. Myron Kayton Kayton Engineering Co. ? 2000 by CRC Press LLC 109.2 Coordinate Frames Navigation is with respect to a coordinate frame of the designer’s choice. Short-range robots navigate with respect to the local terrain or a building’s walls. For navigation over hundreds of kilometers (e.g., automobiles and trucks), various map grids exist whose coordinates can be calculated from latitude-longitude (Fig. 109.1). NATO land vehicles use a Universal Transverse Mercator grid. Long-range aircraft and ships navigate relative to an earth-bound coordinate frame, the most common of which are latitude-longitude-altitude and rectangular x, y, z (Fig. 109.1). The most accurate world-wide reference ellipsoid is described in [WGS-84, 1991]. Spacecraft in orbit around the earth navigate with respect to an earth-centered, inertially nonrotating coordinate frame whose z axis coincides with the polar axis of the earth and whose x axis lies along the equator. Interplanetary spacecraft navigate with respect to a sun-centered, inertially nonrotating coordinate frame whose z axis is perpendicular to the ecliptic and whose x axis points to a convenient star [Battin, 1987]. 109.3 Categories of Navigation Navigation systems can be categorized as: 1. Absolute navigation systems that measure the state vector without regard to the path traveled by the vehicle in the past. These are of two kinds: ?Radio systems (Section 109.5). They consist of a network of transmitters (sometimes also receivers) on the ground or in satellites. A vehicle detects the transmissions and computes its position relative to the known positions of the stations in the navigation coordinate frame. The vehicle’s velocity is measured from the Doppler shift of the transmissions or from a sequence of position measurements. ?Celestial systems (Section 109.6). They measure the elevation and azimuth of celestial bodies relative to the land level and North. Electronic star sensors are used in special-purpose high-altitude aircraft and in spacecraft. Manual celestial navigation was practiced at sea for millennia (see Bowditch). 2. Dead-reckoning navigation systems that derive their state vector from a continuous series of measurements beginning at a known initial position. There are two kinds, those that measure vehicle heading and either FIGURE 109.1 Latitude-longitude-altitude coordinate frame. f = geodetic latitude; OP is normal to the ellipsoid at B; l = geodetic longitude; h = BP = altitude above the reference ellipsoid = altitude above mean sea level. ? 2000 by CRC Press LLC speed or acceleration (Section 109.4) and those that measure emissions from continuous-wave radio stations whose signals create ambiguous “lanes” (Section 109.5). Dead reckoning systems must be reinitialized as errors accumulate and if power is lost. 3. Mapping navigation systems that observe and recognize images of the ground, profiles of altitude, sequences of turns, or external features (Section 109.7). They compare their observations to a stored database, often on compact disc. 109.4 Dead Reckoning The simplest dead-reckoning systems measure vehicle heading and speed, resolve speed into the navigation coordinates, then integrate to obtain position (Fig. 109.3). The oldest heading sensor is the magnetic compass, a magnetized needle or electrically excited toroidal core (called a flux gate), as shown in Fig. 109.2. It measures the direction of the earth’s magnetic field to an accuracy of 2 degrees at a steady velocity below 60-degrees magnetic latitude. The horizontal component of the magnetic field points toward magnetic north. The angle from true to magnetic north is called magnetic variation and is stored in the computers of modern vehicles as a function of position over the region of anticipated travel [Quinn, 1996]. Magnetic deviations caused by iron in the vehicle can exceed 30 degrees and must be compensated in the navigation computer or, in older ships, by placing compensating magnets near the sensor. A more complex heading sensor is the gyrocompass, consisting of a spinning wheel whose axle is constrained to the horizontal plane (often by a pendulum). The ships’ version points north, when properly compensated for vehicle motion, and exhibits errors less than a degree. The aircraft version (more properly called a directional gyroscope) holds any preset heading relative to earth and drifts at 50 deg/hr or more. Inexpensive gyroscopes (some built on silicon chips as vibrating beams with on-chip signal conditioning) are often coupled to magnetic compasses to reduce maneuver-induced errors. The simplest speed-sensor is a wheel odometer that generates electrical pulses. Ships use a dynamic-pressure probe or an electric-field sensor that measures the speed of the hull through the conductive water. Aircraft FIGURE 109.2 Saturated core (“flux-gate”) magnetometer, mounted on a “compass engine” board. The two orthogonal sensing coils (visible) and the drive coil, wound on the toroidal core, measure two components of the magnetic field in the plane of the toroid. (Courtesy of KVH Industries, Inc.) ? 2000 by CRC Press LLC measure the dynamic pressure of the air stream from which they derive airspeed in an air-data computer. The velocity of the wind or sea current must be vectorially added to that of the vehicle, as measured by a dynamic- pressure sensor (Fig. 109.3). Hence, unpredicted wind or current will introduce an error into the dead-reckoning computation. Most sensors are insensitive to the component of airspeed or waterspeed normal to their axis (leeway in a ship, drift in an aircraft). A Doppler radar measures the frequency shift in radar returns from the ground or water below the aircraft, from which speed is inferred. A Doppler sonar measures a ship’s speed relative to the water layer or ocean floor from which the beam reflects. Multibeam Doppler radars or sonars can measure all the components of the vehicle’s velocity. Doppler radars are widely used on military helicopters. The most complex dead-reckoning system is an inertial navigator in which accelerometers measure the vehicle’s acceleration while gyroscopes measure the orientation of the accelerometers. An on-board computer resolves the accelerations into navigation coordinates and integrates them to obtain velocity and position. The gyroscopes and accelerometers are mounted in either of two ways: 1.In servoed gimbals that angularly isolate them from rotations of the vehicle. 2.Fastened directly to the vehicle (“strap-down”), whereupon the sensors are exposed to the maximum angular rates and accelerations of the vehicle (Fig. 109.4). Inertial-quality gyroscopes measure vehicle orientation within 0.1 degree for steering and pointing. Most accelerometers consist of a gram-sized proof-mass mounted on flexure pivots. The newest accelerometers, not yet of inertial grade, are etched into silicon chips. Older gyroscopes contained metal wheels rotating in ball bearings or gas bearings. The newest gyroscopes are evacuated cavities or optical fibers in which counter- rotating laser beams are compared in phase to measure the sensor’s angular velocity relative to inertial space about an axis normal to the plane of the beams. Vibrating hemispheres and rotating vibrating tines are the basis of some navigation-quality gyroscopes (drift rates less than 0.1 deg/h). Fault-tolerant configurations of cleverly oriented redundant gyroscopes and accelerometers (typically four to six) detect and correct sensor failures. Inertial navigators are used aboard naval ships, in airliners, in most military fixed-wing aircraft, in space boosters and entry vehicles, in manned spacecraft, in tanks, and on large mobile artillery pieces. FIGURE 109.3 Geometry of dead reckoning. ? 2000 by CRC Press LLC 109.5 Radio Navigation Scores of radio navigation aids have been invented and many of them have been widely deployed, as summarized in Table 109.1. The most precise is the global positioning system (GPS), a network of 24 satellites and a half-dozen ground stations for monitoring and control. A vehicle derives its three-dimensional position and velocity from ranging signals at 1.575 GHz received from four or more satellites (military users also receive 1.227 GHz). The former Soviet Union deployed a similar system, called GLONASS. GPS offers better than 100-m ranging errors to civil users and 15-m ranging errors to military users. Simple receivers were available for less than $300 in 1997. They are used on highways, in low-rise cities, at sea, in aircraft, and in low-orbit spacecraft. GPS provides continuous worldwide navigation for the first time in history. It will make dead reckoning unnecessary on many vehicles and will reduce the cost of most navigation systems. Figure 109.5 is an artist’s drawing of a GPS Block 2F spacecraft, scheduled for launch in the year 2002. Differential GPS (DGPS) employs one or more ground stations at known locations, that receive GPS signals and transmit measured errors on a radio link to nearby ships and aircraft. DGPS improves accuracy (centimeters for fixed observers) and detects faults in GPS satellites. In 1997, the U.S. was conducting experiments with a nationwide DGPS system of about 25 stations. This Wide Area Augmentation System (WAAS) could eventually replace VORTAC and Category I ILS. A denser network of DGPS stations and GPS-emulating pseudolites, whose stations are located at airports, might replace ILS and MLS (below). In 1997, the cost, accuracy, and reliability of such a Local Area Augmentation System (LAAS) were still being compared to existing landing aids but marine LAAS were in operation for navigation into harbors in North America, the North Sea, and the Baltic Sea. The most widely used marine radio aid in 1997 was Loran-C (see Table 109.1). The 100-kHz signals are usable within 1000 nautical miles (nmi) of a “chain” consisting of three or four stations. Chains cover the United States, parts of western Europe, Japan, Saudi Arabia, and a few other areas. The former Soviet Union has a compatible system called Chaika. The vehicle-borne receiver measures the difference in time of arrival of pulses emitted by two stations, thus locating the vehicle on one branch of a hyperbola. Two or more station pairs give a two-dimensional position fix whose typical accuracy is 0.25 nmi, limited by propagation uncer- tainties over the terrain between the transmitting station and the user. The measurement of 100-microsecond FIGURE 109.4 Inertial reference unit. Two laser gyroscopes (flat discs), an accelerometer, an electrical connector, and three shock mounts are visible. This unit is used in Airbuses and many military aircraft such as the F-18 and Comanche helicopter. (Courtesy of Litton Guidance and Control Systems.) ? 2000 by CRC Press LLC time difference is possible with low-quality clocks in the vehicles. Loran is also used by general aviation aircraft for en-route navigation and for nonprecision approaches to airports (in which the cloud bottoms are more than 200 feet above the runway). Loran service will probably be discontinued at the beginning of the 21st century. The most widely used aircraft radio aid is VORTAC, whose stations offer three services: 1. Analog bearing measurements at 108 to 118 MHz (called VOR). The vehicle compares the phases of a rotating cardioid pattern and an omnidirectional sinusoid emitted by the ground station. 2. Pulse distance measurements (DME) at 1 GHz by measuring the time delay for an aircraft to interrogate a VORTAC station and receive a reply, 3. Tacan bearing information conveyed in the amplitude modulation of the DME replies from the VORTAC stations. Omega is a worldwide radio aid consisting of eight radio stations that emit continuous sine waves at 10 to 13 kHz. Vehicles with precise clocks measure their range to a station by observing the absolute time of reception. Other vehicles measure the range differences between two stations in the form of phase differences between the received sinusoids. Differential Omega creates hyperbolic “lanes” that are 10 to 150 nmi wide. The lanes are indistinguishable from each other by measuring phase; hence the vehicle must count lanes from a point of known position. Errors are about 2 nmi due to radio propagation irregularities. Omega is used by submarines, TABLE 109.1 Worldwide Radio Navigation Aids Frequency Number of Number of Users in 1996 System Hz Band Stations Air Marine Space Land Omega 10–13 kHz VLF 8 15,000 10,000 0 0 Loran-C/Chaika 100 kHz LF 50 120,000 550,000 0 25,000 Decca 70–130 kHz LF 150 2,000 20,000 0 0 Beacons* 200–1600 kHz MF 4000 130,000 500,000 0 0 Instrument Landing { 108–112 MHz 329–335 MHz VHF 1500 150,000 0 0 0 System (ILS)* UHF VOR* 108–118 MHz VHF 1500 180,000 0 0 0 SARSAT/COSPAS 121.5 MHz { 243,406 MHz VHF 5 satellites 200,000 200,000 0 100,000 UHF Transit 150, 400 MHz VHF 7 satellites 0 0 0 0 PLRS 420–450 MHz UHF None 0 0 0 2,000 JTIDS 960–1213 MHz L None 500 0 0 0 DME* 962–1213 MHz L 1500 90,000 0 4 0 Tacan* 962–1213 MHz L 850 15,000 0 4 0 Secondary Surveillance 1030, 1090 MHz L 800 250,000 0 0 0 Radar (SSR)* Identification Friend or Foe (IFF) GPS-GLONASS 1227, 1575 MHz L 24 + 24 satellites 120,000 275,000 4 125,000 Satellite Control { 1760–1850 MHz 2200–2300 MHz S 10 0 0 200 0 Network (SCN) S Spaceflight Tracking { 2025–2150 MHz 2200–2300 MHz S 3 satellites 0 0 50 0 and Data Network 10 ground (STDN) Radar Altimeter 4200 MHz C None 20,000 0 0 0 MLS* 5031–5091 MHz C 30 100 0 0 0 FPQ-6, FPQ-16 radar 5.4–5.9 GHz C 10 0 0 0 0 Weather/map radar 10 GHz X None 10,000 0 0 0 Shuttle rendezvous radar 13.9 GHz Ku None 0 0 4 0 Airborne Doppler radar 13–16 GHz Ku None 20,000 0 0 0 SPN-41 carrier-landing monitor 15 GHz Ku 25 1600 0 0 0 SPN-42/46 carrier-landing radar 33 GHz Ka 25 1600 0 0 0 *Standardized by International Civil Aviation Organization. ? 2000 by CRC Press LLC over-ocean general-aviation aircraft, and a few international air carriers. It was scheduled to be decommissioned in 1997. Landing guidance throughout the western world, and increasingly in China, India, and the former Soviet Union, is with the Instrument Landing System (ILS). Transmitters adjacent to the runway create a horizontal guidance signal near 110 MHz and a vertical guidance signal near 330 MHz. Both signals are modulated such that the nulls intersect along a line in space that leads an aircraft from a distance of about 10 nmi to within 50 ft above the runway. ILS gives no information about where the aircraft is located along the beam except at two or three vertical marker beacons. Most ILS installations are certified to the International Civil Aviation Orga- nization’s (ICAO) Category I, where the pilot must abort the landing if the runway is not visible at an altitude of 200 ft. One hundred ILSs (in 1996) were certified to Category II, which allows the aircraft to descend to 100 ft before aborting for lack of visibility. Category III allows an aircraft to land at still lower weather ceilings. Category III landing aids are of special interest in Western Europe, which has the worst flying weather in the developed world. Category III ILS detects its own failures and switches to a redundant channel within one second to protect aircraft that are flaring-out (within 50 ft of the runway) and can no longer execute a missed approach. Once above the runway, the aircraft’s bottom-mounted radar altimeter measures altitude and either the electronics or the pilot guides the flare maneuver. Landing aids are described by Kayton and Fried [1997]. Throughout the western world, civil aircraft use VOR/DME whereas military aircraft use Tacan/DME for en-route navigation. In the 1990s, China and the successor states to the Soviet Union were installing ICAO- standard navigation aids (VOR, DME, ILS) at their international airports and along the corridors that lead to them from the borders. Overflying western aircraft navigate inertially, with Omega, or with GPS. Domestic flights within the Soviet Union depended on radar tracking, non-directional beacons, and an L-band range- angle system called “RSBN”. They will eventually upgrade to a satellite-based enroute and landing system. U.S. Navy aircraft use a microwave scanning system at 15.6 GHz to land on aircraft carriers; NASA’s space shuttle uses the Navy system to land at its spaceports. Another microwave landing system (MLS) at 5 GHz was supposed to replace the ILS in civil operations, especially for Categories II and III. However, experiments from FIGURE 109.5 Global positioning satellite, Block 2F. (Courtesy of Rockwell.) ? 2000 by CRC Press LLC 1990 to 1997 showed that differential GPS could achieve an accuracy better than 1 m as a landing aid. Hence, it is likely that a LAAS will replace or supplement ILS, which has been guaranteed to remain in service at least until the year 2010 (Federal Radionavigation Plan). NATO may use MLS or a LAAS as a portable landing aid for tactical airstrips. All the space-faring nations operate worldwide radio networks that track spacecraft, compute their state vectors, and predict future state vectors using complex models of gravity, atmospheric drag, and lunisolar perturbations. NASA operates three tracking and data relay satellites (TDRS) that track spacecraft in low earth orbit with accuracies of 10 to 50 m and 0.3 m/s. Specialized ground-based tracking stations monitor and reposition the world’s many communication satellites [Berlin, 1988]. Other specialized stations track and communicate with deep space probes. They achieve accuracies of 30 m and a few centimeters per second, even at enormous inter- planetary distances, due to long periods of observation and precise orbit equations (see [Yuan, 1983]). Position-location and position-reporting systems monitor the state vectors of many vehicles and usually display the data in a control room or dispatch center. Some vehicles derive their state vector from the ranging modulations; others merely report an independently derived position. Table 109.1 lists Secondary Surveillance Radars that receive coded replies from aircraft so they can be identified by human controllers and by collision- avoidance algorithms. The table also lists the U.S. NASA and military spacecraft-tracking networks (STDN and SCN). Tracking and reporting systems have long been in use at marine ports, for airplane traffic control and for space vehicles. They are increasingly being installed in fire trucks, police cars, ambulances, and delivery- truck fleets that report to a control center. The aeronautical bureaucracy calls them Automatic Dependent Surveillance (ADS) systems. The continuous broadcast of on-board-derived position (probably GPS-based) may become the basis of the worldwide air traffic control system of the early 21st century. Several commercial communication satellites plan to offer digital-ranging services worldwide. The intermit- tent nature of commercial fixes would require that vehicles dead-reckon between fixes, perhaps using solid- state inertial instruments. Thus, if taxpayers insist on collecting fees for service, private comm-nav networks may replace the government-funded GPS and air-traffic communication network in the next century. World- wide traffic control over oceans and undeveloped land areas would become possible. Military communication-navigation systems measure the position of air, land, and naval vehicles on battle- fields and report to headquarters; examples are the American Joint Tactical Information Distribution System (JTIDS) and the Position Location Reporting System (PLRS). A worldwide network of SARSAT-COSPAS stations monitors signals from satellite-based transponders lis- tening on 121.5, 243, and 406 MHz, the three international distress frequencies. Software at the listening stations calculates the position of Emergency Location Transmitters within 20 kilometers, based on the observed Doppler-shift history, so that rescue vehicles can be dispatched. Thousands of lives have been saved world- wide, from arctic bush-pilots to tropical fishermen. 109.6 Celestial Navigation Human navigators use sextants to measure the elevation angle of celestial bodies above the visible horizon. The peak elevation angle occurs at local noon or midnight: elev angle (degrees) = 90 – latitude + declination Thus at local noon or midnight, latitude can be calculated by simple arithmetic. Tables of declination, the angle of the sun or star above the earth’s equatorial plane, were part of the ancient navigator’s proprietary lore. The declination of the sun was first publicly tabulated in the fifteenth century in Spain. When time became measurable at sea, with a chronometer in the nineteenth century and by radio in the twentieth century, off- meridian observations of the elevation of two or more celestial bodies were possible at any known time of night (cloud cover permitting). These fixes were hand-calculated using logarithms, then plotted on charts. In the 1930s, hand-held sextants were built that measured the elevation of celestial bodies from an aircraft using a bubble-level reference instead of the horizon. The accuracy of celestial fixes was 3–10 miles at sea and 5–20 miles in the air, limited by the uncertainty in the horizon and the inability to make precise angular measurements on a pitching, rolling vehicle. Kayton (1990) reviews the history of celestial navigation at sea and in the air. ? 2000 by CRC Press LLC The first automatic star trackers were built in the late 1950s. They measured the azimuth and elevation of stars relative to a gyroscopically stabilized platform. Approximate position measurements by dead reckoning allowed the telescope to point within a fraction of a degree of the desired star. Thus, a narrow field-of-view was possible, permitting the telescope and photodetector to track stars in the daytime. An on-board computer stored the right ascension and declination of 20–100 stars and computed the vehicle’s position. Automatic star trackers are used in long-range military aircraft and on space shuttles in conjunction with inertial navigators. Clever design of the optics and of stellar-inertial signal-processing filters achieves accuracies better than 500 ft [Kayton and Fried, 1997]. Spacecraft use the line-of-sight to the sun and stars to measure orientation (for attitude control). Earth- pointing spacecraft usually carry horizon scanners that locate the center of the earth’s carbon-dioxide disc. All spacecraft navigate by radio tracking from earth. When interplanetary spacecraft approach the target planet, the navigation computers (on earth) transform from sun-centered to planet-centered coordinates by observing star occultations and transmitting the images to earth for human interpretation. During the Apollo translunar missions, crews experimentally measured the angle between celestial bodies and the earth or moon with a specially designed manual sextant coupled to a digital computer which calculated the state vector. Other experiments have been made in which American and Soviet crews used manual sextants to observe the angle between celestial bodies and landmarks on earth, from which state vectors were calculated. Autonomous land vehicles on other planets and certain military spacecraft may need celestial navigation. 109.7 Map-Matching Navigation As computer power grows, map-matching navigation is becoming more important. On aircraft, mapping radars and optical sensors present a visual image of the terrain to the crew. Automatic map-matchers have been built, since the 1960s, that correlate the observed image to stored images, choosing the closest match to update the dead-reckoned state vector. More commonly, aircraft and cruise missiles measure the vertical profile of the terrain below the vehicle and match it to a stored profile. Matching profiles, perhaps hourly, reduces the long- term drift of their inertial navigators. The profile of the terrain is measured by subtracting the readings of a baro-inertial altimeter (calibrated for altitude above sea level) and a radar altimeter (measuring terrain clear- ance). An on-board computer calculates the autocorrelation function between the measured profile and each of many stored profiles on possible parallel paths of the vehicle. The on-board inertial navigator usually contains a digital filter that corrects the drift of the azimuth gyroscope as a sequence of fixes is obtained. Hence the direction of flight through the stored map is known, saving the considerable computation time that would be needed to correlate for an unknown azimuth of the flight path. Marine versions profile the seafloor with a sonar and compare the measured profile to stored bottom maps. GPS is adequate for automotive navigation except in high-rise cities, in tunnels, and on streets with heavy foliage. To fill coverage gaps, map-matching software can take advantage of the fact that the vehicle remains on roads. On the highway, dead-reckoning or GPS errors can be rectified to the nearest road. In cities, turns can be correlated with the nearest intersection of matching geometry. An accuracy of several meters is possible if all streets are included on the stored map (e.g., alleys, driveways, and parking garages). The most complex mapping systems observe their surroundings, usually by digitized video, and create their own map of the surrounding terrain. Guidance software then steers the vehicle. In 1997, such systems were in development for hazardous sites such as nuclear plants, waste-disposal facilities, and battlefields, and for unmanned planetary exploration. Delivery robots in buildings are furnished with a map and need only find their successive destinations while avoiding obstacles. They navigate by following stripes on the floor, by observing infrared beacons, or by observing the returns from on-board ultrasonic sonar or laser radar. 109.8 Navigation Software Navigation software is sometimes embedded in a central processor with other avionic-system software, some- times confined to one or more navigation computers. The navigation software contains algorithms and data ? 2000 by CRC Press LLC GPS POSITIONING SYSTEM DELIVERS HIGH ACCURACY system for real-time differential GPS (DGPS) positioning will deliver submeter accuracy to Earth satellites and ground-based users worldwide. Developed at NASA’s Jet Propulsions Laboratory, the system could improve real-time position accuracy to a few decimeters for single-frequency users and 10 cm or better for dual frequency users. In addition to high accuracy, the system provides nearly complete separation of GPS orbit and clock corrections and continuous determination of inter- frequency delay biases for all GPS satellites and reference receivers. Key features include: the use of dynamic orbit estimation, which depends on high-accuracy satellite force models, signal models, geophysical models, and geometric models, in a Kalman filter formulation; use of real-time stochastic estimation to minimize orbit and clock errors arising from quasi-random variations in atmospheric propagation relays and solar radiation pressure; simultaneous processing of smoothed pseudorange and continuous carrier phase data; and use of the stable solar-magnetic reference frame, rather than an Earth-fixed frame, in computing the ionosphere corrections. System operation began in January 1997. Early tests show approximate user differential range errors of less than 20 cm throughout the coverage area, with a North American reference network only. More comprehensive tests with additional global reference sites will be conducted. (Reprinted with permission of NASA Tech Briefs, 20(10), 30, 1996). A ? 2000 by CRC Press LLC that process the measurements made by each sensor (e.g., inertial or air data). It contains calibration constants, initialization sequences, self-test algorithms, reasonability tests, and alternative algorithms for periods when sensors have failed or are not receiving information. In the simplest systems, the state vector is calculated independently from each sensor; most often, the navigation software contains multisensor algorithms that calculate the best estimate of position and velocity from several sensors. Prior to 1970, the best estimate was calculated from a least squares algorithm with constant weighting functions or from a frequency-domain filter with constant coefficients. Now, a Kalman filter calculates the best estimate from mathematical models of the dynamics of each sensor. Digital maps, often stored on compact disc, are carried on some aircraft and land vehicles so position can be visually displayed to the crew. Military aircraft superimpose their navigated position on a stored map of terrain and cultural features to aid in the penetration of and escape from enemy territory. Civil operators had not invested in digital data bases as of 1996. Algorithms for waypoint steering and for control of the vehicle’s attitude are contained in the software of the flight management and flight control subsystems. Specially equipped aircraft (sometimes ships) are often used for the routine calibration of radio navigation aids, speed and velocity sensors, heading sensors, and new algorithms. 109.9 Design Trade-Offs The designers of a navigation system conduct trade-offs for each vehicle to determine which navigation systems to use. Tradeoffs consider the following attributes: ? Cost, including the construction and maintenance of transmitter stations and the purchase of on-board electronics and software. Users are concerned only with the costs of on-board hardware and software. ? Accuracy of position and velocity, which is specified as a circular error probable (CEP, in meters or nautical miles). The maximum allowable CEP is often based on the calculated risk of collision on a typical mission. ? Autonomy, the extent to which the vehicle determines its own position and velocity without external aids. Autonomy is important to certain military vehicles and to civil vehicles operating in areas of inadequate radio-navigation coverage. ? Time delay in calculating position and velocity, caused by computational and sensor delays. ? Geographic coverage. Radio systems operating below 100 kHz can be received beyond line of sight on earth; those operating above 100 MHz are confined to line of sight. On other planets, new navigation aids—perhaps navigation satellites or ground stations—will be installed, ? Automation. The vehicle’s operator (on-board crew or ground controller) receives a direct reading of position, velocity, and equipment status, usually without human intervention. The navigator’s crew station disappeared in aircraft in the 1970s. Human navigators are becoming scarce, even on ships, in the 1990s, because electronic equipment automatically selects stations, calculates waypoint steering, and accommodates failures. Defining Terms Circular Error Probable (CEP): Radius of a circle, centered at the destination, that contains 50% of the navigation measurements from a large sample. Ecliptic: Plane of earth’s orbit around the sun. Inertial Space: Any coordinate frame whose origin is on a freely falling (orbiting) body and whose axes are nonrotating relative to the fixed stars. It is definable within 10 –7 degree/h. Lanes: Hyperbolic bands on the earth’s surface in which continuous-wave radio signals repeat in phase. Nautical Mile (nmi): 1852 m, exactly. Approximately 1 min of arc on the earth’s surface. State vector: Six-component vector, three of whose elements are position and three of whose elements are velocity. Update: The intermittent resetting of the dead-reckoned state vector based on absolute navigation measure- ments (see Section 109.3). Related Topic 102.2 Communications Satellite Systems: Applications References R.H. Battin, An Introduction to the Mathematics and Methods of Astrodynamics, Washington: AIAA Press, 1987, 796 pp. P. Berlin, The Geostationary Applications Satellite, Cambridge: Cambridge University Press, 1988, 214 pp. N. Bowditch, The American Practical Navigator, Washington, D.C.: U.S. Government Printing Office, 1995, 873 pp. M. Kayton, Navigation: Land, Sea, Air, and Space, New York: IEEE Press, 1990, 461 pp. M. Kayton and W.R. Fried, Avionics Navigation Systems, 2nd ed., New York: Wiley, 1997, 773 pp. R.A. Minzner, The U.S. Standard Atmosphere 1976, NOAA Report 76-1562, NASA SP-390, 1976 or latest edition, 227 pp. NASA, Space Network Users Guide, Greenbelt, Md.: Goddard Space Flight Center, 1988 or latest edition, 500 pp. B.W. Parkinson and J.J. Spilker, Eds., Global Positioning System, Theory and Applications, American Institute of Aeronautics and Astronautics, 1996, 1300 pp., 2 vols. J. Quinn, “1995 revision of joint U.S./U.K. geomagnetic field models,” J. Geomagnetism and Geo-Electricity, 1996. U.S. Air Force, NAVSTAR-GPS Interface Control Document, Annapolis, Md.: ARINC Research, 1991, 115 pp. U.S. Government, Federal Radionavigation Plan, Department of Transportation, 1996, 229 pp., issued biennially WGS-84, U.S. Defense Mapping Agency, World Geodetic System 1984, Washington, D.C.: 1991. J. Yuen, Deep Space Telecommunication Systems Engineering, New York: Plenum Press, 1983, 603 pp. Y. Zhao, Vehicle Location and Navigation Systems, Massachusetts: Artech House, 1997, 345 pp. ? 2000 by CRC Press LLC Further Information IEEE Transactions on Aerospace and Electronic Systems, bimonthly through 1991, now quarterly. Proceedings of the IEEE Position Location and Navigation Symposium (PLANS), biennially. Navigation, journal of the U.S. Institute of Navigation, quarterly. Journal of Navigation, Royal Institute of Navigation (UK), quarterly. AIAA Journal of Guidance and Control, bimonthly. Commercial aeronautical standards produced by International Civil Aviation Organization (ICAO, Montreal), Aeronautical Radio, Inc. (ARINC, Annapolis, Md.), Radio Technical Commission for Aeronautics (RTCA, Inc., Washington) and European Organization for Civil Aviation Equipment (EUROCAE, Paris). ? 2000 by CRC Press LLC