Efficient task scheduling for performance optimization

This chapter presents a typical fog network consisting of multiple fog nodes (FNs), wherein some task nodes (TNs) have heavy computation tasks, while some helper nodes (HNs) have spare resources for sharing with their neighboring nodes. To minimize the delay of every task in such a fog network, it formulates a noncooperative game called paired offloading of multiple tasks (POMT) to model the competition among TNs for the communication resources and computation capabilities of HNs. Further, the chapter describes the distributed POMT task scheduling algorithm. It also presents a novel energy-efficient fog computing framework, which exploits collaboration among the user device nodes, but each node can have heterogeneous computing capacity and network quality. The chapter also describes a low complexity maximal energy-efficient task scheduling algorithm for achieving energy-efficient task scheduling among FNs, and extends the jointly optimal management of computation and communication resources for computation offloading.