Observation scheduling and simulation in a global telescope network

• Published in: Future generation computer systems 2019-06, Vol.95, p.116-125
• Main Authors: López-Casado, Carmen, Pérez-del-Pulgar, Carlos, Muñoz, Víctor F., Castro-Tirado, Alberto J.
• Format: Article
• Language: English
• Subjects: Astronomy; Multi-objective optimization; Scheduling; Simulation environment; Telescope network
• ISSN: 0167-739X; 1872-7115
• DOI: 10.1016/j.future.2018.12.066

The GLObal Robotic-telescope Array is an e-infrastructure composed of a network of telescopes with the aim of providing citizen science capabilities. To allow it, the network is managed by a scheduler that receives observation requests from users, and decides the best telescope to execute them. The objective is to maximize the number of accepted observations and minimize the elapsed time between the user request and its execution. This issue arises as a multi-objective optimization problem that can be solved by means of different methods. Therefore, the aim of this work is to develop a new probabilistic algorithm that decides the best telescopes to execute a requested observation, taking into consideration the optimization problem. To perform a comparison of the new algorithm with others, a model of the telescope network has also been created and validated through information obtained from the real network. Finally, a comparative of the new algorithm with previously developed ones has been carried out to demonstrate their performance in the model. •Simulation framework for a telescope network based on discrete-event system.•Multi-objective optimization problem: Maximize acceptance and minimize time.•A new telescope decision algorithm based on a generalized linear regression model.•Pareto frontier comparative among different decision algorithm.•The new algorithm shows the best performance.