v3.0.4 / chapter 1 of 2 / 01 sep 08 / greg goebel / public domain
* The US was a pioneer in the development of navigation satellites, introducing the "Transit" system in the 1960s, mostly for targeting submarine-launched long-range missiles. Transit was effective but required expensive receiver systems, and so the US moved to the "Global Positioning System (GPS)" or "Navstar" system in the 1980s. This chapter provides a description of the Transit and GPS systems.
* After the Soviets had launched the first artificial Earth satellite, "Sputnik 1", in 1957, some researchers realized that radio transmissions from a satellite in a well-defined orbit could be used to determine the position of a receiving station back on the Earth.
The initial approach towards using a satellite for position location was based on measuring the "Doppler shift" of the satellite's radio transmissions as it passed overheard. A Doppler shift is a change in the frequency of a signal due to the motion of the platform emitting the signal, with the frequency increasing as the platform approaches and decreasing as the platform moves away. The pattern of Doppler shifts of the satellite would give the receiving station its location relative to the satellite's orbital track. If the path and timing of the satellite's orbit were precisely known, the Doppler shifts would establish the location of the receiver station.
This approach required complicated electronic equipment, as well as measurements from several orbital passes of the satellite to get an accurate position fix. Doppler positioning was the basis for the first satellite positioning system, known as "Transit", which was introduced by the US military in the 1960s, mainly for use by US Navy ballistic missile submarines. It was designed by the US Navy and the Applied Physics Laboratory (APL) of Johns Hopkins University in Maryland.
The Transit system was based on a constellation of six satellites in circular 90-degree polar orbits at an altitude of about 1,000 kilometers (620 miles), with three ground stations to control the satellites, and satellite receiver systems carried on submarines and other large naval vessels; the Navy went so far as to build a custom computer, the AN/UYK-1, to support the Transit system on submarines, since no stock computer could fit through the hatches and had the required durability. Only three of the satellites were actually used for positioning, with the other three set aside as on-orbit spares. The Transit system was operated by the US Naval Space Operations Center, with headquarters at Point Mugu, California.
The operational Transit satellites transmitted on two frequencies, 149.99 megahertz (MHz) and 399.97 MHz. An Earth receiver measured the Doppler shift of the frequencies and also downloaded the satellite's position coordinates, broadcast by the satellite itself every two minutes. While receivers could obtain a position fix using only one frequency, much higher accuracy could be obtained by measuring both frequencies, since errors caused by variations in the atmosphere affected the two frequencies differently, allowing the errors to be averaged out. Locations of naval vessels could be determined to an error radius of 80 to 100 meters (260 to 330 feet), adequate for targeting nuclear-armed missiles, but locations of fixed sites could be determined to less than 20 meters (66 feet) using repeated measurements.
* The history of the Transit launches is a bit confusing. The following
table gives a summary, with the table followed by a more detailed
explanation:
date designation launch vehicle notes
_________________________________________________________________________
17 sep 59 Transit 1A Thor Able II Launch failure.
13 apr 60 Transit 1B* Thor Able-Star
22 jun 60 Transit 2A* Thor Able-Star
30 nov 60 Transit 3A* Thor Able-Star Launch failure.
22 feb 61 Transit 3B* Thor Able-Star
29 jun 61 Transit 4A* Thor Able-Star
15 nov 61 Transit 4B* Thor Able-Star
18 dec 62 Transit 5A Scout X-3 Operational prototype.
05 apr 63 Transit 5A-2 Scout X-3 Launch failure.
15 jun 63 Transit 5A-3 (7) Scout X-3 Partial failure.
28 sep 63 Transit 5BN-1* Thor Able-Star Partial failure.
05 dec 63 Transit 5BN-2* Thor Able-Star 1st operational Transit.
21 apr 64 Transit 5BN-3 Thor Able-Star Launch failure.
04 jun 64 Transit 5C (9) Scout X-4
06 oct 64 Transit-O 1 (5B-4)* Thor Able-Star Failed in orbit.
13 dec 64 Transit-O 2 (5B-5)* Thor Able-Star Failed in orbit.
11 mar 65 Transit-O 3* Thor Able-Star
24 jun 65 Transit-O 4 (5B-6)* Thor Able-Star
13 aug 65 Transit-O 5 (5B-7)* Thor Able-Star Failed in orbit.
22 dec 65 Transit-O 6 (10) Scout A
28 jan 66 Transit-O 7 (11) Scout A
25 mar 66 Transit-O 8 (12) Scout A
19 may 66 Transit-O 9 (13) Scout A
18 aug 66 Transit-O 10 (14) Scout A
13 apr 67 Transit-O 12 (15) Scout A
18 may 67 Transit-O 13 (16) Scout A
25 sep 67 Transit-O 14 (17) Scout A
02 mar 68 Transit-O 18 Scout A First RCA Transit.
27 aug 70 Transit-O 19 Scout A
02 sep 72 TIP 1 Scout B Technology test.
30 oct 73 Transit-O 20 Scout A
12 oct 75 TIP 2 Scout D Improved Transit-O.
01 oct 76 TIP 3 Scout D
28 oct 77 Transit-O 11 Scout D SATRACK payload.
15 may 81 Nova 1 Scout G Production TIP.
12 oct 84 Nova 3 Scout G
03 aug 85 Transit-O 24,30 Scout G First dual launch.
16 sep 87 Transit-O 27,29 Scout G
26 apr 88 Transit-O 23,32 Scout G
16 jun 88 Nova 2 Scout G
25 aug 88 Transit-O 25,31 Scout G
_________________________________________________________________________
(*) Indicates other payloads in launch. The parenthesis indicate an
alternate designation, for example "Transit-O 14" was also known as
"Transit 17".
The initial shot was on 17 September 1959, when "Transit 1A", a demonstrator
satellite, was launched from Cape Canaveral on a Thor Able II booster. It
didn't make orbit. It was followed by six more or less successful shots of
demonstrator satellites, beginning with "Transit 1B" on 13 April 1960 and
following through "Transit 4B" on 15 November 1961. They were all launched
by Thor Able-Star boosters from Cape Canaveral, with each shot carrying a
number of payloads. Some of the payloads were secret, such as the "SolRad"
satellites, which were labelled as solar radiation studies spacecraft but
which were actually military signals intelligence spy satellites.
The Transit 1 through Transit 3 satellites were in the form of a sphere with a rim around the middle, with some variation between the three series, while the Transit 4 satellites were in the shape of drums. The Transit 4 series were nuclear powered, each carrying a "radioisotope thermoelectric generator (RTG)" to generate electricity.
After the demonstrator satellites validated the concept, work began on operational satellites, all of which were launched from Vandenberg Air Force Base in California. They were intended to be launched by the Scout light booster, which meant they had to be smaller than the demonstrators, and they also had to be ruggedized since the Scout gave a rougher ride than the Thor Able-Star.
There were seven shots of prototypes of operational satellites, beginning with "Transit 5A" on 18 December 1962 and ending with "Transit 5C" on 4 June 1964. The Transit 5BN series were atomic powered, but the decision was made to go with solar power for the operational satellites, since it was cheaper and less bureaucratically troublesome. Some of the launches were with Thor Able-Stars, and the launches often included other payloads.
The operational satellites were known as "Transit-O", of course with "O" for "Operational". Since the military phonetic alphabet gives "Oscar" for "O", they were also known as "Transit-Oscars" or just "Oscars". They had nothing to do with the long-running amateur radio satellites known as "OSCARs"; the "Transit-O" designation is used here to reduce confusion. The Transit-O spacecraft used a gravity-gradient stabilization system, weighed 65 kilograms (143 pounds), and had an octagonal bus surrounded by four paddle-like solar arrays, with a long rod extending from the center to provide the gravity gradient stabilization.
Initial Transit-O shots, beginning with "Transit-O 1" on 6 October 1964, were disappointing, with the satellites quickly failing in orbit. These spacecraft were built directly by the Navy; APL then took up manufacturing the satellites and reliability began to improve. Manufacturing was finally contracted out to RCA through APL, with the first RCA-built satellite, "Transit-O 18", launched on 2 March 1968. Reliability would continue to improve until a satellite had a typical lifetime measured in decades.
A number of the Transit satellites were launched along with "Transit 5E" spacecraft, which were not navigation satellites, instead being small multipurpose science research satellites. Records tend to be contradictory on the Transit 5Es and they are not discussed in detail here. There are stories that some Transit-O launches were actually of auroral studies satellites nicknamed "OWL", but this appears to be a garbling of the historical records.
An experimental technology-validation satellite for the "Transit Improvement Program (TIP)" was launched on 2 September 1972, being designated "TIP 1". It led to the "TIP" and "TIP 3" prototypes, which led in turn to the operational "Nova" series of Transit satellites, with three launched in the 1980s. The Novas had the same general configuration as the Transit-O spacecraft, but they were more sophisticated and much bigger, with a weight of about 160 kilograms (353 pounds).
Transit-Os continued to be launched, using an uprated Scout G booster that could put two into orbit at one time, to the final Transit launch on 25 August 1988. The Transit system actually remained in service until 31 December 1996. Such satellites as are still operational remain in use for ionospheric studies.
* Even before the first Transit satellite was in orbit, researchers were considering a better approach that would eventually result in the modern Global Positioning System.
Suppose an entire constellation of navigation satellites was launched into precisely-known orbits. If a receiver could determine the exact position of several of these satellites at the same time, then the receiver could precisely determine its own position through triangulation.
Determining the position of the satellites was the trick. Ground-based radar tracking was dismissed as impractical, since it required too much equipment and would give away the position of the ground station. A subtler scheme was adopted instead: the satellites would emit signals with distinctive, precisely timed pulse patterns that a receiver could read to determine the distance of each satellite. The receiver had to be able to identify different satellites from the pulse patterns they emitted.
Design of the pulse patterns was based on a technique that had been devised by astronomers for radar studies of the Moon and nearby planets. Radar operates by transmitting a radio pulse and measuring the time it takes the pulse to travel to a target, be reflected, then travel back to the original transmitter. Since sending a radar pulse to the Moon or a nearby planet involved a long time delay, the radar pulse had to have a distinctive pattern to ensure that the exact timing of pulse transmission and reflection could be determined. To accomplish this, astronomers used a set of pulses emitted in a "pseudo-random noise (PRN)" sequence. The pulse pattern appeared to be random noise that never repeated itself, at least over the length of time of interest, but was generated by a predictable algorithm. The reflected pulse could be compared against the original pulse pattern using a mathematical procedure known as "correlation" to determine the exact timing.
The engineers working on the navigation satellites like the idea of using PRN sequences, though instead of transmitting an on-off pulse pattern, the satellites would send the pulse pattern using a continuous signal that varied between two frequencies, an approach known as "frequency shift keying (FSK)". Furthermore, the PRN sequences could be designed so that each navigation satellite had its own unique and non-overlapping sequence; in formal terms, the sequences were said to be "orthogonal". This allowed all the satellites to transmit on the same frequency and not be confused by a receiver. A receiver then only had to pick up a single frequency to "hear" all the satellites, reducing complexity and cost.
* For this scheme to give precise locations, the clocks on the satellite and on the receiver had to be precisely synchronized. A timing error of no more than a microsecond would result in an error of about 300 meters (1,000 feet). While the satellites could carry precision clocks, such clocks would be far too clumsy and expensive for the Earth-based receivers. However, there turned out to be a way to build a receiver that could provide accurate position data using a clock no more accurate than a typical cheap digital wristwatch.
Suppose the receiver's clock is used to help determine the distances to four satellites. Due to clock inaccuracy, these distances will be inexact, and so are known as "pseudo-ranges". The receiver's position could be thought of as being at the intersection of four invisible spheres, with the radius of each defined by the receiver at one end and a satellite at the other. The pseudo-ranges could give an estimate for the length of all four of these radii, but due to the clock error the invisible spheres would not intersect. It is, however, straightforward arithmetic to then use the intersection error to compute the clock error, subtract it, and make the spheres link up.
* The last major issue was the altitude of the GPS satellite constellation. Satellites are generally either placed in low Earth orbit, a few hundred kilometers high, or in geostationary orbit over the equator, 36,000 kilometers (22,360 miles) high, where they take 24 hours to orbit the Earth and remain in a fixed position relative to the Earth as it turns under them.
Putting the satellites into low Earth orbit would reduce the size and cost of the boosters required to launch them, and would also reduce the power required for the transmitters on the satellites. However, obtaining adequate coverage would demand a large number of satellites. Putting them into geosynchronous orbit would reduce the number of satellites, but it would require more powerful launchers and transmitters, and it would not provide good coverage of the polar regions. The altitude finally chosen was a compromise: a circular orbit with an altitude of 20,200 kilometers (12,545 miles) and a period of 12 hours. At that altitude, 17 satellites would be enough to make sure that four of them, the minimum number needed to establish a position, would always be visible from any location on the Earth's surface.
The GPS constellation finally implemented actually has 24 satellites, consisting of 21 in active operation, plus three spares. The 24 satellites operate in six different orbital "planes" (an orbital path shared by multiple satellites), with four satellites in each plane. The planes are inclined 55 degrees with respect to the equator. The GPS satellites are also fitted with nuclear blast detectors as a secondary mission, replacing the early "Vela" nuclear blast surveillance satellites in this role.
* The table below lists GPS satellite launches:
date spacecraft booster notes
_____________________________________________________________________
31 may 67 Timation 1 Thor Agena D
30 sep 69 Timation 2 LT Thor Agena D
14 jul 74 Timation 3 Atlas Centaur
23 jun 77 NTS-2 Atlas Centaur
22 feb 78 GPS-1 1 Atlas Centaur
13 may 78 GPS-1 2 Atlas Centaur
07 oct 78 GPS-1 3 Atlas Centaur
11 dec 78 GPS-1 4 Atlas Centaur
09 feb 80 GPS-1 5 Atlas Centaur
26 apr 80 GPS-1 6 Atlas Centaur
19 dec 81 GPS-1 7 Atlas Centaur failure
14 jul 83 GPS-1 8 Atlas Centaur
13 jun 84 GPS-1 9 Atlas Centaur
08 sep 84 GPS-1 10 Atlas Centaur
09 oct 85 GPS-1 11 Atlas Centaur
14 feb 89 GPS-2 1 Delta 2
10 jun 89 GPS-2 2 Delta 2
18 aug 89 GPS-2 3 Delta 2
21 oct 89 GPS-2 4 Delta 2
11 dec 89 GPS-2 5 Delta 2
24 jan 90 GPS-2 6 Delta 2
26 mar 90 GPS-2 7 Delta 2
02 aug 90 GPS-2 8 Delta 2
01 oct 90 GPS-2 9 Delta 2
26 nov 90 GPS-2A 10 Delta 2
07 jul 91 GPS-2A 11* Delta 2
23 feb 92 GPS-2A 12 Delta 2
10 apr 92 GPS-2A 13 Delta 2
07 jul 92 GPS-2A 14 Delta 2
09 sep 92 GPS-2A 15 Delta 2
21 nov 92 GPS-2A 16 Delta 2
18 dec 92 GPS-2A 17 Delta 2
03 feb 93 GPS-2A 18 Delta 2
30 mar 93 GPS-2A 19* Delta 2
13 may 93 GPS-2A 20 Delta 2
26 jun 93 GPS-2A 21 Delta 2
30 aug 93 GPS-2A 22 Delta 2
26 oct 93 GPS-2A 23 Delta 2
10 mar 94 GPS-2A 24* Delta 2
28 mar 96 GPS-2A 25 Delta 2
16 jul 96 GPS-2A 26 Delta 2
12 sep 96 GPS-2A 27 Delta 2
10 jan 97 GPS-2R 1 Delta 2 failure
23 jul 97 GPS-2R 2 Delta 2
05 nov 97 GPS-2A 28 Delta 2
07 oct 99 GPS-2R 3 Delta 2
11 may 00 GPS-2R 4 Delta 2
16 jul 00 GPS-2R 5 Delta 2
10 nov 00 GPS-2R 6 Delta 2
30 jan 01 GPS-2R 7 Delta 2
29 jan 03 GPS-2R 8* Delta 2
31 mar 03 GPS-2R 9 Delta 2
21 dec 03 GPS-2R 10 Delta 2
20 mar 04 GPS-2R 11 Delta 2
23 jun 04 GPS-2R 12 Delta 2
06 nov 04 GPS-2R 13 Delta 2
24 sep 05 GPS-2R M1 Delta 2 AKA GPS-2R 14
25 sep 06 GPS-2R M2 Delta 2 AKA GPS-2R 15
17 nov 06 GPS-2R M3 Delta 2 AKA GPS-2R 16
17 oct 07 GPS-2R M4 Delta 2 AKA GPS-2R 17
20 dec 07 GPS-2R M5 Delta 2 AKA GPS-2R 18
15 mar 08 GPS-2R M6 Delta 2 AKA GPS-2R 19
_____________________________________________________________________
(*) Indicates other payloads in launch.
The three "Timation" satellites were technology demonstrators to validate GPS
concepts, leading to the "NTS-2" satellite in 1977, which was a prototype for
a functional GPS satellite. It led to the launch of ten (not counting a
failure) "Block 1" GPS or "Navstar" satellites, built by Rockwell
International. This constellation was also a prototype, not intended for
full operational service, and these spacecraft differed from the later
operational GPS satellites in that they were placed into orbits with an
inclination of 63 degrees, not 55 degrees.
The first "Block 2" operational satellite was launched in 1989, and was followed by 8 more in that series into 1990. The next 15 were slightly improved "Block 2A" satellites, launched into 1996. The full 24-satellite operational constellation was finally completed with the launch of GPS-2A 24 on 10 March 1994. Both the Block 2 and Block 2A satellites were also built by Rockwell. The Block 2/2A satellites have a design lifetime of over 7 years.
A series of 13 (not counting a failure) "Block 2R" satellites was launched into 2004. They were built by Lockheed Martin. Current launches are of the "Block 2R-M" ("M" for "Modernized") specification, with additional GPS signals. The first Block 2R-M launch was in 2005, with a total of eight in the series planned. The Block 2R / 2R-M satellites will be followed by at least 12 further improved "Block 2F" satellites to be built by Boeing, which bought Rockwell's space assets in 1996. The USAF is now performing studies of a follow-on "GPS 3" or "Block 3" system, with the first Block 3 satellites to be launched in 2009.
Ground system improvements will be implemented as part of the Block 2F program. The ground system includes a number of elements. Overall GPS direction resides at the "GPS Master Control Station (GPS MCS)" located at Falcon Air Force Base, outside Colorado Springs, Colorado. The MCS is linked to four remote active-tracking ground antenna stations and five passive-tracking monitor stations. The ground stations, which are at precisely-known locations, forwards GPS satellite broadcasts to the GPS MCS. The GPS MCS measures the timing of the signals, and then uploads any necessary corrections.
* The US military, having designed GPS to support their operations, wanted to make sure that they were the only ones entitled to the full accuracy of the system, the error being no more than 10 meters (33 feet). They introduced "noise" or "dithering" into the signals transmitted by the satellites, with the error coarsened for civilian users to no more than 100 meters (330 feet). The "noise" is apparently generated as a classified code pattern that military GPS receivers can filter out. The highly accurate military service is known as the "precise positioning service", while the degraded civilian service is known as the "standard positioning service". The military refers to this scheme as "selective availability (SA)".
Selective availability proved controversial. Civilian users felt that the value of precise positioning was great enough for public use that it was wrong-headed to deny it. What made the argument even more troublesome was that there were ways to get around selective availability. In 1980, MIT researchers demonstrated a method of greatly reducing the uncertainty in non-military GPS positioning. Since they knew the satellite orbits with precision, then if they had a ground receiver whose exact position was known by other means, they could then measure the distances to the satellites using the coded signals, and calculate the difference between the true distance and the distance given by the coded signals. The corrections could then be broadcast locally to GPS receivers in the area to allow them to correct their own positions accordingly.
This scheme became known as "differential GPS", and allowed cheap GPS receivers to obtain locations to within about 10 meters (33 feet), with the aid of a error correction signal. The availability of differential GPS, in the view of many civilian GPS enthusiasts, made selective availability a joke. What made the joke even more ironic was that some US government organizations implemented GPS error-correction broadcast networks, also known as "GPS augmentation services", that were accessible by civilians. In particular, the US Coast Guard established a "National Differential GPS (NDGPS)" network that originally provided differential GPS error correction signals in coastal areas, but has been extended, partly with help of the US Department of Transportation, to nearly all of the United States.
* The military was still reluctant to give up selective availability, but it was finally turned off by executive order on 2 May 2000. Measurements of GPS accuracy performed by the US National Oceanic & Atmospheric Administration (NOAA) showed that before selective availability was turned off, 95% of the position readings sampled fell within a 45 meter radius, and then zoomed to a 6.3 meter radius.
Interestingly, even with selective availability disabled, differential GPS provides such high accuracy that the military is using it as well. Military differential GPS systems are being developed under a US Air Force GPS accuracy improvement initiative, which also involves distribution of more accurate data for GPS satellite orbits and other GPS parameters. The goal is to improve accuracies to less than 5 meters (16 feet).
The US military is working on "local denial" techniques to prevent adversaries from making use of GPS in a combat theater. Details are classified, but observers suspect local denial will involve some type of selective jamming from an aircraft or ground station, with US military GPS receivers able to operate in the presence of such jamming. Contingency plans have also been drawn up to allow partly or completely disabling the system if a warning is received of a terrorist attack using some class of GPS-aided weapons.
* All GPS satellites up to and including the Block 2R satellites broadcast two microwave carrier channels, with timing based on two rubidium and two cesium atomic clocks. The first carrier is at the "L1" frequency, 1,575.42 MHz, and the second is at the "L2" frequency, 1,227.60 MHz. The two carriers provide somewhat different sets of signals:
The civilian L1 signal is known as the "coarse acquisition (C/A)" signal. This signal carries a 1,023-bit PRN code, which as mentioned earlier uniquely identifies a particular GPS satellite.
The L1 and L2 military signals are both based on "precision (P)" PRN codes, about 6 x 10^12 bits long, with a cycle time of a week. Since delays in the propagation of radio waves through the atmosphere change more or less predictably with frequency, the use of a P signal on each carrier allow military receivers to provide some compensation for such delays. During military operations, the P codes can be encrypted by another level of coding, known as a "Y code", to prevent an adversary from trying to "spoof" GPS receivers with bogus GPS signals. The two P codes are combined with the C/A signal to provide high-resolution position data.
* The Block 2R-M satellites add a new military or "M" code to both carrier frequencies, and also provide a new L2 code for civilians designated "L2C". The dual M codes give increased resistance to jamming by using "spread spectrum" techniques, and also are believed to provide a capability to deny an enemy use of the GPS signal, though details are classified. The second civilian signal gives civilian users increased ability to compensate for atmospheric delays. The Block 2F satellites will add a third carrier signal designated "L5" at 1,176.45 MHz. The new L5 signal is intended for civilian applications in air traffic control.
Current thinking about GPS 3 envisions that it will provide 100 times greater signal power, mostly through the use of spot beams, and allow the targeting of GPS-guided munitions to less than a meter. GPS 3 is expected to reduce the number of orbital planes from six to three, using nonrecurring orbits. The Air Force also wants to improve the reliability and security of GPS.
* The guts of a GPS receiver consists of three functional blocks:
A simple GPS receiver can only pick up a single GPS signal at one time, obtaining the multiple signals needed to get a position fix on an alternating or "multiplexed" (interleaved) basis. A more sophisticated receiver may have five "channels", allowing it to pick up five satellites at a time. Five channels are required, even though only four satellites are necessary for a fix, because one satellite may drop below the horizon during the operation, requiring acquisition of another. Some high-end GPS receivers actually have a dedicated channel for each satellite in the entire constellation.
The internal electronics are enclosed in a protective case. The case also contains batteries and access to external power, along with a keyboard and display, and may possibly include digital interfaces to allow the receiver to be hooked up to a computer.
Early single-channel military receivers were big and heavy, weighing about 9 kilograms (20 pounds), but modern GPS receivers are light and compact. GPS chipsets are available from a number of manufacturers, and are also sold in some cases as complete modules that can be interfaced into an electronic system. Inexpensive cellphones are now available with GPS capability.
* As GPS usage and accuracy increases, concerns over sources of error has increased as well. There are three primary sources of error:
Incidentally, the accuracy of a GPS position fix is also partly dependent on the positions of the visible satellites in the sky. Position fixes are about two or three times more accurate if the satellites are scattered all over the sky than they are if they are clustered close together.
* Since the GPS satellites carry precision time references, they can be used to provide timing information accurate to within 100 nanoseconds of the Universal Time Coordinated (UTC) atomic clock. The gear required to obtain this accurate timing is much more expensive than a standard GPS receiver, costing thousands of dollars, but it's still much cheaper than an atomic clock.
* One of the peculiarities of the GPS satellites is that they operate on a clock that "rolls over" to zero every 1,024 weeks, with the calendar beginning on at 00:00 hours (midnight) on 5 January 1980. This "week number roll-over (WNRO)" was part of the spec for satellite operation, and so should not have been a surprise to anyone building a GPS receiver.
However, as the year 2000 ("Y2K") approached, there were concerns about glitches in the computing infrastructure from computing hardware and software that hadn't been designed properly to handle Y2K "rollover", with some even predicting a global catastrophe. GPS was seen as a particularly worrisome case, since the very first WNRO was scheduled to occur just before midnight on Saturday, 21 August 1999. The uncertainly that older GPS receivers could properly handle the discontinuity sent manufacturers into a frenzy of testing and updates. In any case, the GPS network passed over both the WRNO and Y2K hurdles with no serious difficulties. Y2K arrived with no major problems and the issue has now been generally forgotten.
* Although GPS had been used by US forces in a limited fashion during the "tanker war" in the Persian Gulf in the late 1980s and during the US invasion of Panama in 1989, the military effectiveness of GPS was dramatically proven during the Persian Gulf war in 1990:1991. GPS was used in its planned roles to guide bombers to targets, allow infantry and armored units to locate their locations in the featureless desert, and position artillery for precise fire. It was also used to guide experimental missiles to their targets.
GPS was a great success, even though the constellation was not fully operational. When the crisis started in August 1990, only 14 GPS satellites were in orbit. Two more were launched and put into service in record time, with 16 satellites in service when the ground war began. This was enough to give military forces continuous two-dimensional positions, but only intermittent three-dimensional positions.
There was also a lack of military-qualified GPS receivers. Only 4,000 were available when the crisis began, and so the military simply ordered thousands of commercial handheld GPS receivers. Antennas had to be improvised to allow the use of the receivers inside sealed-off armored vehicles. Ironically, to make use of the civilian GPS receivers, the military had to turn off selective-availability coding; the critics made much of this inconsistency. Saddam Hussein's forces did not exploit this opportunity to use GPS in their own operations.
* The US military has expanded their use of GPS since the Gulf War, acquiring bombs, missiles, and even artillery shells with GPS or differential GPS guidance. GPS guidance allows such weapons to accurately strike targets in any weather, day or night. GPS-guided weapons were used extensively during the Balkans bombing campaign in spring 1999, the Afghanistan campaign in the winter of 2001:2002, and the invasion of Iraq in the spring of 2003.
Ideally, targets can be acquired and located by the launch aircraft or other platform using imaging radar or other sensors, with target coordinates downloaded immediately into the weapon using a hardwired connection or infrared link. The weapon is then launched and guides itself to the target coordinates without further operator intervention. Ground forces can also now use laser target designator / rangefinder units to obtain the GPS coordinates of a target and relay them automatically by radio to a strike platform. A GPS guidance system is much cheaper than most other types of weapons guidance system, though for greater accuracy a weapon may also be fitted a "terminal seeker", such as a laser seeker, or an optical or infrared camera, for pinpoint targeting.
The US military is applying GPS in other imaginative ways. For example, cargo parasails have been developed that can be dropped at high altitude by a transport aircraft, and then sail to a remote location, automatically guided by a GPS receiver, allowing the transport to remain out of range of air defenses near the landing zone.
* Originally, the military simply regarded GPS as a navigational aid. Few thought of say, fitting a GPS receiver to an artillery shell, and so the problem of jamming wasn't seriously considered. The GPS signals are highly vulnerable to jamming as they are extremely weak, providing about the equivalent energy as a household light bulb thousands of kilometers away, a billion times weaker than the signals picked up by a broadcast television set.
GPS can be effectively jammed with a brute-force "broadband" jammer that throws out radio noise all over the spectrum. A specialized GPS jammer that selectively operates on GPS frequencies would be even more effective, and both Russian companies and American academic institutions have developed such jammers. GPS signals have actually been jammed by accident on a number of occasions, interfering with the navigation systems of aircraft.
There is no shortage of antijamming ideas. Of course, GPS-guided smart munitions always have a backup inertial navigation system to take over when the GPS signal has been demonstrably compromised, though the accuracy of the INS is poorer than that of GPS. Due to the troublesome press over the potential of GPS jamming, manufacturers of smart munitions insist that their weapons are actually inertially guided and are merely "GPS-aided". Weapons can also be designed to have a "home-on-jam" capability to attack the GPS jammer. However, trading a GPS-guided munition that costs thousands of dollars at minimum for a $500 USD jammer would not be a bargain.
GPS receivers can improve their resistance to jamming by improving the selectivity of their reception. One approach is to use multiple separated antennas so that the angle of the signal being received can be determined, assisting in the rejection of signals coming from the ground and not the sky. Increasing GPS satellite power output would help, and in fact the GPS 3 satellites now under consideration may use focused spot beams to ensure much higher signal power in specific combat theaters.
The US Defense Advanced Research Projects Agency (DARPA), which performs research on advanced military technologies, has performed studies on a concept in which a network of ground stations and high-flying, long-endurance UAVs could produce high-power location signals that could be used by standard GPS receivers. DARPA refers to the ground stations and UAVs as "GPX pseudolites". Current status of this research is unclear.
* The first civilian application of GPS was on large ships, where the relatively high expense of the early GPS receivers was not such a problem. Gradually, prices fell to the point where they became common on smaller vessels; GPS receivers are now a common option on cars, and as mentioned it possible to buy cellphones with integrated GPS receivers. A GPS-enabled cellphone is particularly useful, since it links communications with positioning -- this allows, for example, parents to determine the location of their children carrying a GPS cellphone.
Interesting civilian applications of GPS under consideration or being implemented include a flight-data recorder, or aircraft "black box", that tracks the position of an aircraft over time; robotic earth excavation; disposal of toxic substances; monitoring of suspension bridges to warn of impending bridge failure; and space capsules for scientific and commercial experiments that would, after reentry, deploy a parasail and glide to a predetermined landing site for recovery.
Geophysicists have been exploiting GPS since the mid-1980s, using it to measure continental drift and the movement of the Earth's surface in geologically active regions. They have been able to obtain accurate surface measurements to within a few millimeters through a procedure known as "carrier tracking", which is even more accurate than differential GPS. Carrier tracking actually senses the phase of the carrier signals on which the location code sequences are broadcast. It is, not surprisingly, a tricky and subtle procedure, and not applicable for general use.
A particularly interesting potential scientific application of GPS is in observations of changes in the ionosphere, the ionized layer of the upper atmosphere from 80 to 600 kilometers (37 to 370 miles), through a procedure known as "radio occultation", long used by interplanetary probes. All it consists of is tracking changes in the probe's radio signal as it passes behind another planet, in order to obtain information about the planet's atmospheric density and other parameters. Radio occultation experiments in Earth orbit would involve the launch of satellites carrying GPS receivers. Once in space, ground controllers would observe the timing shifts in the precise GPS signals as the sensing satellites fell under the horizon from the line of sight to a GPS satellite.
A radio occultation experiment built by the US National Aeronautics & Space Administration's Jet Propulsion Laboratory (NASA JPL) was put into orbit in July 2000 on board the German "Challenging Minisatellte Payload (CHAMP)" spacecraft for Earth studies. JPL's "Blackjack" package carried on CHAMP featured a rearward-facing GPS antenna to perform the radio occultation experiments, and also featured a downward-facing antenna to pick up GPS reflections from the ocean surface. The downward-facing antenna was highly experimental, with researchers using it to see if GPS signals could be used to determine ocean surface heights and wave conditions. In principle, the heights could be determined from the time of signal travel, and wave conditions from the spreading of frequencies and travel times by choppy seas, a procedure known as "scatterometry".
The Blackjack package also included a top-mounted GPS antenna for fixing its own position. An improved follow-on to the Blackjack package named "TurboRogue", built by JPL with help from Italy, was flown on the Argentine "Satelite de Aplicationes Cientificas-C (SAC-C)" satellite, launched in November 2001.
* One of the major emerging uses of GPS is for air-traffic control (ATC) systems. Civil air traffic is now becoming increasingly dependent on GPS. Up to the mid-1950s, air traffic control in the US was based on controllers using radio communications and handwritten notes to direct traffic. After a disastrous mid-air collision in 1956, the US Federal Aviation Administration (FAA) set up a system of radars and computers to keep closer track of airline traffic.
Up to the introduction of GPS, long-range civil aircraft navigation was handled by radio beacons on land and aircraft-based inertial navigation systems over oceans. On approach to the runway, an aircraft was directed by radio-based "VOR / DME" radio beacon system to ensure that the aircraft was on the proper approach path. Large airports used the automated "Instrument Landing Systems (ILS)" to then bring aircraft down safely in day or night, in any weather.
GPS has made long-distance navigation much simpler, eliminating the reliance on ground-based navigational beacons. However, unaugmented GPS does not have the resolution needed for approach and landing systems, and so the US Federal Aviation Administration (FAA) has implemented a GPS "Wide Area Augmentation System (WAAS)", which reduces airliner position errors from a hundred meters (330 feet) with unaugmented GPS to less than 3 meters (10 meters), and also provides vertical position information.
WAAS is based on a network of 25 ground stations at precisely-known positions around the US. These stations do not directly transmit error signals to airliners. Instead, they pick up GPS signals, determine errors, and transmit the error data to one of two master stations. The error correction data is then transmitted to communications satellites in geostationary orbit, and relayed in turn to aircraft with WAAS gear. WAAS also provides integrity monitoring, with WAAS receivers providing an indicator to alert a pilot when the system is not working properly. Interestingly, the error correction data is transmitted by the communications satellites in the same frequency band as the GPS signals themselves.
WAAS went into initial service in the summer of 2003, though functionality was restricted at the outset and of course few aircraft had WAAS gear at the time. WAAS is now being adopted by Canada, Mexico, Japan, and India; many other countries are likely to follow.
Since WAAS doesn't have the accuracy for blind landings in bad weather, GPS augmented systems have been considered, with the objective of building a cheaper replacement for the traditional Instrument Landing System that would feature GPS augmentation transmitters located at airports. The FAA has backed the "Local Area Augmentation Systems (LAAS)", but LAAS has run into difficulties and was on hold at last notice. Honeywell Corporation, the prime contractor for LAAS, has worked to help set up a "Ground Based Augmentation System (GBAS)" and an associated "Ground-based Regional Augmentation System (GRAS)" in Australia, and it may become the basis for the final GPS-based approach system.
Another element of GPS-aided air navigation is "Automatic Dependent Surveillance B (ADS-B)", in which aircraft carry gear to transmit their GPS location coordinates to ATC centers, allowing the aircraft to be tracked without use of radar. Ground systems are now being developed that integrate all the ADS-B tracks to give air traffic controllers a map of flight traffic around them. Pilots of aircraft with an ADS-B "in" capability will be able to spot other aircraft on their own displays.
A flight control network based on ADS-B instead of radar is not only cheaper, with an ADS-B station a third the cost of a radar system and much easier to maintain, but provides better coverage and is much more accurate, with positions given in terms of meters instead of hundreds of meters. Once the system becomes widespread, it will be possible to safely reduce minimum required aircraft separations in crowded airspace. In 2005, the FAA decided to implement a national ADS-B network over the next decade, installing at least 400 ADS-B ground stations, and retiring 200 redundant radar stations as the system becomes established. Australia is planning on setting up an ADS-B network in 2006.
Yet another GPS tool for airliners now under development is the "Traffic Collision Avoidance System IV (TCAS-IV)", which will use ADS-B to obtain precise locations of airliners and determine if they are on a collision course. The current "TCAS-III" system uses radar.
Such widespread use of GPS for air-traffic control would mean that the abrupt failure of the GPS network might have disastrous consequences. As a result, WAAS also includes "integrity services" that provide notification to airliners through the communications satellites if the GPS network goes down.
In addition, the US Air Force and Navy are developing an augmented GPS approach and landing control system for military aircraft, known as the "Joint Precision Approach & Landing System (JPALS)", which will be compatible with the US civilian WAAS and LAAS systems. "Hands-off" aircraft carrier landings have been performed by pilots in fighter jets using prototype JPALS technologies, and JPALS promises to be very useful for controlling the new generation of "unmanned aerial vehicles (UAVs)" now in development.
* GPS program officials are justly proud of the accomplishments of the system. In 2003 the US GPS industry had revenues of $17 billion USD, and in terms of the tax return to the government it more than pays its own way. Said one official: "This is the only Defense Department program I know of that has a positive return on investment." However, not surprisingly the economic benefits of GPS do not directly translate into proportional funding for the Air Force, and the USAF has complained about bearing the full costs of the system. Broadening the funding mechanism would require a major change in the system's charter and setting up a US national program office.
