Computing Offloading with Edge Layering Mechanism for Mobile Ad-Hoc Networks

Mobile ad-hoc networks (MANETs) encounter significant challenges in supporting efficient edge computing due to their dynamic topologies, resource limitations, and constrained bandwidth. To address these challenges, this paper proposes a novel edge layering mechanism (ELM) combined with a deep reinforcement learning (DRL)-based computing offloading algorithm. ELM organizes edge computing nodes into hierarchical layers to enhance the efficient utilization of computing resources. The proposed offloading algorithm formulates the problem as a partially observable Markov decision process (POMDP) and employs a multi-agent deep reinforcement learning (MADRL) approach to optimize task offloading decisions across these layered edge nodes. Simulation results demonstrate the effectiveness of the proposed method in enhancing task execution efficiency, reducing system latency, and improving energy utilization compared to baseline algorithms. This work provides valuable insights into leveraging edge layering and DRL for scalable and adaptive edge computing in MANET environments.