An Improved Dependent Task Offloading in Device-Edge-Cloud using Favorite Predecessor Task

Cloud computing offers fast processing and large storage, while edge computing reduces delay and local device resource issues. A proper offloading strategy is needed among local devices, edge servers, and cloud server to properly utilize the resources. Especially for the dependent task offloading needs extra attention during offloading as result of one task has an impact on other tasks. In device-edge-cloud collaboration, edge selection has a great impact on dependent task offloading. In this paper, we proposed an approach with favorite predecessor (FP) information during edge server selection in a device-edge-cloud environment. This FP-based edge selection strategy reduces the communication delay between the dependent tasks. For this reason, the proposed edge selection approach shows a greater improvement (around 24%) in the overall task completion time than the greedy approach.