Adaptive Hierarchical Offloading for Mobile Edge Computing in High-Mobility SWIPT-Enabled Networks

Mobile devices are affected by limited energy and channel resources in the dynamic scenarios with high mobility and complexity, which will trigger the risks such as task failure or low offloading efficiency. This article proposes a multiuser and multiserver MEC network framework based on simultaneous wireless information and power transfer (a.k.a., SWIPT). We first consider multiple mobile devices with fixed task information in which the tasks can be either processed in-local or offloaded to the MEC server for processing via uplink transmission. Our method also embraces a hierarchical demand-weighted index (HDWI) and priority channel transmission scheduling rule, which can evaluate the status of device services and effectively conduct the hierarchical offloading of tasks. In this case, our model not only ensures the continuity of the service provided by mobile devices but also evaluates the relationship between energy consumption and device delay during the offloading process. Finally, we propose an efficient mobile device cost hierarchical offloading algorithm (MCHOA) to deal with the issues produced by the constructed multiobjective optimization mathematical model. MCHOA complies with the principle of HDWI and is combined with a multiobjective evolutionary algorithm based on decomposition to solve various mathematical tasks including obtaining the Pareto optimal curve regarding the average time consumption and the average energy consumption of devices. The experimental results show that MCHOA can simultaneously reduce time consumption and energy consumption costs by at least 13.3% and 37.5%, respectively. Our experiments also validate the superiority of the proposed algorithm and the application prospect of the model.