Distributed optimization for task offloading in hybrid edge networks: integrating D2D and LEO satellite resources

This paper investigates computational offloading in integrated edge networks combining device-to-device (D2D) collaboration and low Earth orbit (LEO) satellite assistance, aiming to maximize the global quality-of-experience (QoE) for ground user devices (GUDs) under joint resource and coverage constraints. We first establish the NP-hardness of this problem. To address this complexity, we reformulate the problem as a QoE-oriented offloading game (QOUO-Game), demonstrating its equivalence to a potential game with guaranteed Nash equilibrium existence. A distributed QOUO algorithm is proposed, enabling autonomous decision-making among GUDs through iterative resource arbitration and interference-aware scheduling. The algorithm ensures convergence to equilibrium strategies while balancing workloads across terrestrial and orbital resources. Comprehensive simulations validate the QOUO algorithm’s superiority. This work advances distributed optimization in next-generation edge networks through theoretical guarantees and practical coordination mechanisms.