Energy-Efficient Topological Dependency and Data-Aware Splittable Task Offloading in Mobile Edge Networks

Rapid advancements in Vehicle-to-Everything (V2X) and Mobile Edge Computing (MEC) have posed significant challenges for Mobile Devices (MDs) in managing complex tasks and addressing mobility effects. While MDs employ task offloading to mitigate resource constraints, the presence of task topological dependencies and mobility limitations diminishes its effectiveness, thereby impacting energy efficiency and service quality. In response, we propose the Energy-Efficient Topological Dependency and Data-aware Splittable Task Offloading (ETDS) framework. ETDS categorizes task data into stateless and stateful segments, enabling the offloading of stateless data, independent of topology constraints. Furthermore, ETDS optimizes task offloading timing and target locations, capitalizing on opportunistic offloading due to MD mobility, consequently reducing energy consumption associated with task offloading transmissions. Simulation results reveal that ETDS can significantly reduce MDs energy consumption by 30% to 63% when compared to traditional dependent task offloading schemes, demonstrating consistent performance across different MD speeds and real-world parameter configurations.