The Development of the Global Positioning System (GPS)
Personal navigation devices and smart phones incorporate GPS. This helps their users in finding their destination and supports numerous location dependent services. One billion of such devices are in use today. Every airplane and ship carries several GPS receivers. The traffic density in both areas could not be maintained without GPS. Additionally, GPS is starting to be used operationally for critical maneuvers such as landing aircraft. Farming machines drive to sub-decimeter accuracy using precise point position and the drift of continental plates is tracked with an accuracy of millimeters per year. Finally, GPS plays a crucial role in the synchronization of communication systems and power grids. A disruption of GPS would have severe consequences on the daily life of every citizen.
GPS is the “Global Positioning System” developed by Col. Prof. Bradford Parkinson – today’s recipient of the Eduard Rhein Prize.
Although hardly imaginable, there was a world without GPS and even without satellites. On October 4th, 1957 Sputnik was launched and started orbiting the earth at a velocity near 30.000 km/h. This created a significant Doppler shift on its radio signal, which was observed by several groups of researchers. They used the time dependency of that frequency shift to determine Sputnik’s orbit. Frank McClure had the ingenious idea to reverse the setting and to determine the position of the user under the assumption of a known orbit. TRANSIT – the first satellite navigation system was born. It had four to seven satellites at a height of roughly 1000 km above ground. A receiver, which could predict its own relative movement during roughly 15 minutes, was able to determine its absolute position on earth. The requirement of predicting one’s own relative movement was met on submarines using sophisticated inertial platforms as well as in geodesy using stationary receivers. The two dimensional position had an accuracy of roughly 25 meters. In geodesy, a number of measures brought the accuracy to the meter level. This capability led to the implementation of a first worldwide geodetic reference system. The most prominent correction to our previous knowledge was the shift of Hawai by more than one kilometer.
However, TRANSIT had its limitations: not every mobile user wanted to carry a high precision inertial platform, nor was everyone patient enough to wait 100 minutes to see the next satellite pass, and aircraft were seeking for the third dimension, as well. Something had to happen. This was recognized by Ivan Getting. He initiated work at Aerospace Corporation for the U.S. Air Force. A first study by James Woodford and Hideyoshi Nakamura proposed to use atomic clocks on the satellite for creating a stable time reference, and to determine the user’s position by measuring the phase of the signals from four satellites. In parallel, the navy also had started work in a similar direction.
Col. Bradford Parkinson was appointed director of the Air Force program in November 1972 and of a merged joint program in 1973. He assembled a team of experts from leading universities such as MIT and Stanford with relevant experience. He also contracted Aerospace Corporation for selective support. Over the Labor Day weekend in September 1973, Parkinson convened a meeting with a core team of experts in the Pentagon with the purpose of defining the system. Translated into today’s language and non-military terms, the requirements were to develop a satellite navigation system to determine positions in three dimensions, instantly, everywhere, at any time, in a scalable way with inexpensive user equipment (less than 10.000 USD). Technological breakthroughs had been made recently; first atomic clocks had been built, first satellites had been launched into the harsh space environment of geosynchronous orbits, first integrated circuits had been developed, and spread spectrum communication had started to be used in military systems. They were crucial for this endeavor at the limit of the 1970s capabilities.
The design of Parkinson’s team included 21 satellites (later 24) that would carry atomic clocks and fly at a high altitude (ca. 20.200 km). Signals from at least four satellites were to be capture at a time to match the three degrees of freedom for position and an additional one for the time offset. Robert Gold had just developed his family of spreading codes, which was ideally suited for broadcasting simultaneously spread spectrum signals from several satellites on the same frequency. Gold codes were adopted for the open signal, and are used in every smart phone today. They represented the ideal solution for minimizing inter-satellite biases in the receiver, since all signals were passing the same analog front-end. A second encrypted signal would be transmitted on two frequencies for estimating ionospheric delay. It was more wideband and meant to be evaluated by military receivers only. A key decision in signal design was to transmit all codes as well as the carrier in a phase-coherent manner. This would later become the enabler for carrier phase positioning and the associated millimeter accuracies. A low-rate navigation message was finally modulated on the ranging signal to provide all necessary information to receivers. The satellites were planned in six orbital planes with a period of 11 h 58 minutes to allow for testing, and an inclination of 62 degrees (later changed to 55 degrees) to achieve a good geometry for position determination by the receiver.
Additionally, a number of questions relating to the launcher, power supply and testing were addressed from the start by the team working under the guidance of Col. Parkinson. Although the location of the design meeting in September 1973 at the premises of the Pentagon and the program itself were both military, the system was conceived for dual use from the beginning. Parkinson presented and defended this proposal, and obtained approval to launch the development program on December 14th 1973. Maintaining this approval for such a costly program over the necessary period of time was not an easy task. It required constant attention and effort. The real work started in early 1974 with a team of around 30 engineers in the program, and another 25 at Aerospace Corporation.
The Munich-based company Efratom had built a small atomic clock at that time. A variant of this clock was incorporated on the satellite NTS-1, which was a dowry from a previous program of the navy. The Efratom concept inspired a second line of development, using Cesium atoms contracted to Rockwell International. The resulting clock was flown on the first GPS satellite in February 1978. Besides reaction wheels, radiation hardened clocks were the most critical element in satellite longevity and one of the biggest challenges in the program. The mean time before failure had to be around 6-8 years for economically sustaining a constellation of 24 satellites. The satellites 3-10 already achieved an average life-time of 9.1 years. This was in sharp contrast to GLONASS, which started with 3 years, and saw its constellation decay to very few satellites in the early phase of the Russian Federation. The long life-time of GPS satellites was the result of a careful choice of redundancies, a rigorous specification and selection of parts, and of a detailed analysis of every single failure.
GPS was a military system, which meant that its monitoring stations could only be placed in locations that were under control of the USA. Orbits had, therefore, to be predicted over long periods of time. A number of phenomena became apparent in such predictions: earth tides and pole wander moved the reference stations, irregularities in the gravitational field shook the satellites in their orbit, the radiation pressure pushed them around, and relativity influenced the pace of time due to the relative movement and gravity. Additionally, imperfections of the equipment such as clock-offsets, inter-frequency biases, phase center variations and the like caused uncertainties, and had to be captured. Finally, ionospheric and tropospheric delays, and multipath influenced the measurements as well. All of this had to be taken into account during orbit determination and was solved in the context of the program. In more recent times, the International GPS Service (IGS) has setup a network of several hundred monitoring stations, and developed the capability to determine the orbits of GPS satellites with an accuracy of 2 centimeters. This accuracy is achieved for satellites flying with roughly 14.500 km/h at a distance of more than 20.000 km, based on signals received with a power of one tenth of a femtowatt (10-16W).
Col. Prof. Bradford Parkinson has led the team, which developed this system at the frontier of the technology of its time. His intuition, inspiration, and guidance were crucial to the success of GPS. He proved the same diligence later, in his function as the program manager of Gravity Probe B, a satellite designed to verify two predictions of Einstein’s relativity theory: the geodetic effect and Lense-Thirring frame dragging – another project at the limits of feasibility. Parkinson furthermore contributed to numerous techniques related to the use of GPS, including differential positioning, receiver autonomous integrity monitoring, attitude determination, ambiguity resolution, the estimation of ionospheric delay, the mitigation of multipath, and many more.
Today, we would like to honor him for the development of GPS – the system that measures the three missing dimensions in our four dimensional world, and which has become the model for other navigation systems, including Galileo. In its first twenty years, GPS has significantly changed our world, and the pace of change is still increasing.
Prof. Dr. Christoph Günther,
German Aerospace Center,