224
4 Navigations from Ground to Space
WWV could yield a superb reference for a microwave spectrometer. Furthermore,
the Soviet designers had set the Sputnik-1 frequency about 1 kHz from an exact
20 MHz so that any receiver would produce an audio tone of 1 kHz plus or minus the
Doppler shift generated by the satellite motion. This offset of the satellite’s frequency
ensured that the received audio tone never went through zero, varying from about
1500 Hz to about 500 Hz, clearly audible throughout an entire pass.
Weiffenbach was responsible for receiving the signals transmitted by the Sputnik1 while Guier was responsible for the work of the signal processing. By extracting
the Doppler shifts, Guier could estimate the orbital parameters of the Sputnik-1. The
Sputnik-1 stopped transmitting signals on October 26, 1957. Right after, the Sputnik2 was launched by the USSR on November 3, 1957. Weiffenbach and Guier also
received the dual-frequency signals from the Sputnik-2 and used the dual-frequency
Doppler shift data to determine the satellite’s orbit parameters. From the culmination
of this study using the archived Sputnik-1 data and some of the Sputnik-2 data added,
it was demonstrated that a complete set of orbit parameters for a near-earth satellite
could be inferred to useful accuracy from a single set of Doppler shift data. The
demonstration established that the whole Doppler curve was needed, nearly from
the horizon to horizon, and that a first-order correction for ionosphere refraction was
required, as well as an inferred correction for the satellite’s oscillator frequency and
frequency drift. As a result, there was the total number of nine unknown parameters,
where the six is the orbital parameters, the one is the ionosphere correction parameter
and the other two are, respectively, the bias and drift correction parameters of the
satellite’s oscillator frequency. Included in this demonstration was a single-parameter
model of the ionosphere electron density, which was also inferred along with the
six orbital parameters. The measurement equations were established by using the
Doppler shifts of a complete observation pass, and thus the nine unknown parameters
could be estimated with the method of least squares.
On March 17, 1958, Frank T. McClure (1916−1973), who was then the director
of the APL’s Research Center, called Weiffenbach and Guier to his office. After
confirming that their study on determining an approximate orbit from a single pass
of Doppler data had not been exaggerated, McClure presented an inverse solution
problem to them, determining the station’s position while assuming the orbit is
known. This became the well-known “navigation satellite problem”, later. “Go
do an error analysis and let me know the answer ASAP.” said McClure [7]. Clearly,
the number of the unknown parameters needed to be solved for the problem had been
reduced to five, including two plane coordinate parameters of a ground observation
station and three system parameters. The study quickly evolved to the assumptions
that the satellite is cooperative and emits the signals at two frequencies that are very
stable and sufficiently high to effectively eliminate ionosphere refraction errors. And
thus, the number of the system parameters was reduced to only one, the onboard clock
bias. In other words, finally, there were only three unknown parameters needed to be
solved. From the simulation results, it was indicated that the location coordinates of
the ground stations could reach high accuracy, which was exciting and even unbelievable for Weiffenbach and Guier. When they reported excitedly the results back to
McClure, he, of course with his consistently penetrating insight, was not surprised.
4 Navigations from Ground to Space
WWV could yield a superb reference for a microwave spectrometer. Furthermore,
the Soviet designers had set the Sputnik-1 frequency about 1 kHz from an exact
20 MHz so that any receiver would produce an audio tone of 1 kHz plus or minus the
Doppler shift generated by the satellite motion. This offset of the satellite’s frequency
ensured that the received audio tone never went through zero, varying from about
1500 Hz to about 500 Hz, clearly audible throughout an entire pass.
Weiffenbach was responsible for receiving the signals transmitted by the Sputnik1 while Guier was responsible for the work of the signal processing. By extracting
the Doppler shifts, Guier could estimate the orbital parameters of the Sputnik-1. The
Sputnik-1 stopped transmitting signals on October 26, 1957. Right after, the Sputnik2 was launched by the USSR on November 3, 1957. Weiffenbach and Guier also
received the dual-frequency signals from the Sputnik-2 and used the dual-frequency
Doppler shift data to determine the satellite’s orbit parameters. From the culmination
of this study using the archived Sputnik-1 data and some of the Sputnik-2 data added,
it was demonstrated that a complete set of orbit parameters for a near-earth satellite
could be inferred to useful accuracy from a single set of Doppler shift data. The
demonstration established that the whole Doppler curve was needed, nearly from
the horizon to horizon, and that a first-order correction for ionosphere refraction was
required, as well as an inferred correction for the satellite’s oscillator frequency and
frequency drift. As a result, there was the total number of nine unknown parameters,
where the six is the orbital parameters, the one is the ionosphere correction parameter
and the other two are, respectively, the bias and drift correction parameters of the
satellite’s oscillator frequency. Included in this demonstration was a single-parameter
model of the ionosphere electron density, which was also inferred along with the
six orbital parameters. The measurement equations were established by using the
Doppler shifts of a complete observation pass, and thus the nine unknown parameters
could be estimated with the method of least squares.
On March 17, 1958, Frank T. McClure (1916−1973), who was then the director
of the APL’s Research Center, called Weiffenbach and Guier to his office. After
confirming that their study on determining an approximate orbit from a single pass
of Doppler data had not been exaggerated, McClure presented an inverse solution
problem to them, determining the station’s position while assuming the orbit is
known. This became the well-known “navigation satellite problem”, later. “Go
do an error analysis and let me know the answer ASAP.” said McClure [7]. Clearly,
the number of the unknown parameters needed to be solved for the problem had been
reduced to five, including two plane coordinate parameters of a ground observation
station and three system parameters. The study quickly evolved to the assumptions
that the satellite is cooperative and emits the signals at two frequencies that are very
stable and sufficiently high to effectively eliminate ionosphere refraction errors. And
thus, the number of the system parameters was reduced to only one, the onboard clock
bias. In other words, finally, there were only three unknown parameters needed to be
solved. From the simulation results, it was indicated that the location coordinates of
the ground stations could reach high accuracy, which was exciting and even unbelievable for Weiffenbach and Guier. When they reported excitedly the results back to
McClure, he, of course with his consistently penetrating insight, was not surprised.