Autonomous Satellite System Synchronization Schemes via Optical Two-Way Time Transfer and Distributed Composite Clock

To improve the provision of a global satellite navigation service, the German Aerospace Center (DLR) - Institute of Communication and Navigation - is proposing a next-generation global navigation satellite architecture named Kepler. Autonomous synchronization at picosecond-level is a fundamental component of the Kepler concept, achieved via two-way time transfer (TWTT) schemes enabled by optical inter-satellite links (OISLs). This level of synchronization is only achievable if relativistic effects are adequately considered. In this paper we present the synchronization scheme for Kepler: all satellites perform pairwise relativistic TWTT, providing relative clock offsets in a predefined coordinate time scale. These are then distributed across the whole constellation and are used as input for a space-based distributed clock ensemble. Each satellite realizes a local copy of the Kepler system time (KST) by steering a local oscillator, so that all satellites will tend to “beat” the same time, thus achieving a tight synchronization. We show how measurement noise impacts the final synchronization level, in two different designs of the Kepler architecture. Additionally, the impact of constant biases on the system time generation is analyzed. Finally, we assess the impact of the choice of constellation's measurement topologies (open versus closed rings). The synchronization performance is expressed in terms of maximum time offset between any two satellites of the constellation.