Kepler – Our Proposal for a Future Satellite Navigation System

Positioning in the second generation satellite systems GPS, GLONASS, Galileo and Beidou is based on pseudorange measurements, which requires that the orbits and the clock offsets of all satellites are accurately estimated by control centers. Today, the estimation of the latter parameters is achieved using highly stable satellite clocks and an extensive network of ground reference stations. Satellite clocks are a limiting factor in three respects: firstly, they are the most fragile component and limit the lifetime of the satellites. Secondly, their stability is critical to the accuracy with which orbits and clocks can be separated and also to the maximum time lap between updates of the navigation messages. Finally, atomic clocks are prone to rare exceptional events, such as frequency jumps. Under certain conditions such events are threats that need to be addressed in expensive augmentation systems. Ground networks are another limiting factor to current systems. GPS had to expand its original network by NIMA stations, Galileo included a large number of ground stations, and GLONASS as well as Beidou had and are still struggling with the extent of their networks. The costs of maintaining the network of Galileo reference stations are substantial and Brexit is causing worries about the coverage in the South Atlantic. Kepler circumvents these problems. It is a system without atomic clocks and with a ground network that could be reduced to a single station. Furthermore, initial simulations of the GFZ-Potsdam indicate that the signal in space accuracy could be improved by more than a factor of 10. Kepler achieves satellite synchronization by using cavity stabilized lasers as time references as well as two-way inter-satellite laser links. Furthermore, the observability of Kepler navigation signals is tremendously improved by placing the references stations on LEO satellites above the atmosphere. The synchronization to UTC is achieved either from the ground or by placing a small number of ultra-stable clocks on selected satellites. The use of LEO satellites leads to the most cost effective upgrade capability but is more limited by the knowledge of the gravitational field. The Kepler system consists of 24 MEO satellites, which reuse the current orbital positions of the Galileo satellites, and another 4 additional LEO satellites. All satellites carry cavity stabilized lasers, which have an Allan deviation of roughly 10 - 15 for up to 10 seconds. These lasers are synchronized by two - way optical links, which have an Allen deviation that is much smaller. In order to keep optical transceiver terminals simple, the associated links only connect MEO satellites within one orbital plane. A third MEO terminal is oriented more freely to any LEO satellites, in order to connect the respective MEO plane with a LEO satellite. Each LEO satellite carries a number of terminals to simultaneously connect to at least two orbital MEO planes. The LEO satellites thus act as synchronization relays between the orbital planes. The accumulated delay to propagate information from an arbitrary satellite to any other one is below one second, i.e. well in the range of stability of the cavity stabilized lasers. Using spread spectrum optical signals and the optical carrier itself leads to a synchronized constellation with very little uncertainty by combining all time and frequency offsets in a multi-layer composite clock algorithm. The laser links are additionally used for data communications at a data rate of 50 Mbps and to measure inter-satellite ranges with an accuracy of a few tens of a micrometer. The accuracy is thus not limited by measurements but rather by vibration modes of the satellite. A frequency comb transforms the synchronized optical signals into a radio frequency signal of equal stability, which is used to drive generators for the modulation of the optical signals, as well as for the L-band navigation signals transmitted towards earth by a nadir pointing antenna. The latter signals are compatible with the latest specification of Galileo. They are observed by Kepler’s LEO satellites using zenith pointing antennas. Since the latter satellites are flying above the ionosphere, these navigation signals only experience distortions due to the transmit and receive equipment, which allows for a characterization of the associated antenna patterns and biases. All measurements performed on LEO and MEO satellites are easily transported throughout the constellation and processed in one or several places to obtain orbits, biases, and the mentioned time and frequency offsets. The clock corrections and biases are corrected directly during the generation of the navigation signals. The orbital parameters and antenna patterns on the other hand are integrated into the navigation messages. In order to prevent a drift of the satellite based orbits with respect to an earth fixed reference frame at least one ground receiver need to be included. This is confirmed by extensive simulations carried out by GFZ-Potsdam. For robustness, a small number of receivers will typically be used. There are no a priori requirements on their location. Cavity stabilized lasers are robust: lasers have been characterized for space and cavities are made of low expansion glass. Frequency combs have been characterized for space and one was flown on a sounding rocket. Optical transceiver terminals have been built and flown in LEO and GEO orbits. Thus the system described so far seems suited for a long life-time in a space environment. The MEO space environment is, however, even more demanding than the GEO environment, for example, and thus a careful qualification program needs to be carried out in a pre-development program. The approach discussed in this abstract is attractive to other regions of the world as well, in particular with respect to its low requirements on ground infrastructures. Europe can and should take the lead on the approach.      «

Positioning in the second generation satellite systems GPS, GLONASS, Galileo and Beidou is based on pseudorange measurements, which requires that the orbits and the clock offsets of all satellites are accurately estimated by control centers. Today, the estimation of the latter parameters is achieved using highly stable satellite clocks and an extensive network of ground reference stations. Satellite clocks are a limiting factor in three respects: firstly, they are the most fragile component and...     »