Resource Probing-Based Task Offloading in Hierarchical Multi-Hop Edge Computing Networks

While multi-hop task offloading can alleviate the limitations of single-hop schemes in infrastructure-free scenarios, most existing methods depend on global state information, neglect directional searches for the next hop, and ignore resource probing cost. Accordingly, we propose a Resource Probing based Task Offloading (RPTO) solution in hierarchical multi-hop edge computing networks, enabling directional searches while balancing probing cost and service performance. We then formulate RPTO as a joint optimization problem to minimize service cost and design a Deep Deterministic Policy Gradient (DDPG)-based algorithm to solve it. Simulation results demonstrate that the proposed algorithm outperforms baselines, reducing service cost by 67%.