Submitted: February 2023
Abstract
Satellites as means of globally providing internet access are becoming more widespread, with services like Starlink getting more adoption and new services like IRIS²2 being planned. In order to offer high reliability and good coverage great constellations consisting of many satellites are deployed. Typically, these satellites form a Delay-Tolerant-Network (DTN) to exchange information. Some common algorithms used for routing through such networks require knowledge about the network a priori, namely, whether and when satellites are in contact. For big constellations, this can be difficult to compute for a whole constellation.
In this thesis, several methods for optimizations will be discussed, implemented and evaluated. Ideas for the optimizations range from improving the implementation to making use of constellations’ common characteristics and reasoning from them how contact detection could be improved.
As a result, a software application was written that is able to output traces of constellation’s satellites and to efficiently detect and output contacts of constellation’s satellites. Some optimizations yielded notable speed-ups that help reduce the time needed for the process of contact detection.