CELESTE aims to develop a holistic framework to address the problem of resource orchestration in 3D networks, delivering efficient and resilient algorithm designs and solution outcomes, while explicitly accounting for system uncertainties. To this end, CELESTE targets three key dimensions: (a) modeling stochastic and time-varying uncertainties in 3D networks; (b) designing efficient distributed resource orchestration algorithms that remain resilient despite incomplete and imperfect information at distributed network entities; and (c) identifying efficient and resilient operating points for the large-scale distributed networks under various types of uncertainties. The CELESTE framework will integrate theoretical foundations (e.g., stochastic game theory), distributed learning techniques (e.g., reinforcement learning, no-regret learning), and resilient solution concepts (e.g., correlated equilibria) to holistically address joint radio and computing resource orchestration in 3D networks. This integrated approach aimes to conclude uncertainty-aware solutions that are both realistic and computationally-efficient. Representative 6G use case scenarios will be considered for the performance evaluation and validation of the proposed framework.
Funding Agency: Hellenic Foundation for Research and Innovation (H.F.R.I.)
Duration: 02/02/26 – 01/02/28







