GT-PPO: Graph Attention-Based and Sequence-Aware Deep Reinforcement Learning for Adaptive SFC Orchestration in SAGIN-MEC
GT-PPO is proposed, a deep reinforcement learning (DRL)-based approach for online SFC orchestration designed to maximize network profit while minimizing end-to-end (E2E) delay and capturing global dependencies among virtual network functions (VNFs).
Guang-Yu Bian, Jing Wu, Hao Li et al.
· Electronics · 0 citations