END-TO-END SYSTEM PERFORMANCE SIMULATION: A DATA-CENTRIC APPROACH

摘要

In the early times of space industry, the feasibilityof Earth observation missions was directly drivenby what could be achieved by the satellite. It wasclear to everyone that the ground segment wouldbe able to deal with the small amount of data sentby the payload. Over the years, the amounts ofdata processed by the spacecrafts have beenincreasing drastically, leading to put more andmore constraints on the ground segmentperformances - and in particular on timeliness.Nowadays, many space systems require high datathroughputs and short response times, withinformation coming from multiple sources andinvolving complex algorithms. It has becomenecessary to perform thorough end-to-endanalyses of the full system in order to optimise itscost and efficiency, but even sometimes to assessthe feasibility of the mission.This paper presents a novel framework developedby Astrium Satellites in order to meet these needsof timeliness evaluation and optimisation. Thisframework, named ETOS (for “End-to-endTimeliness Optimisation of Space systems”),provides a modelling process with associatedtools, models and GUIs. These are integratedthanks to a common data model and suitableadapters, with the aim of building suitable spacesystems simulators of the full end-to-end chain.A big challenge of such environment is to integrateheterogeneous tools (each one being well-adaptedto part of the chain) into a relevant timelinesssimulation.