Machine Learning for Context-Aware Cross-Layer Optimization

This chapter presents a predictive scheduling model and develops the predictive multi-tier operations scheduling (PMOS) algorithm, where the fog control node is assumed to be aware of users’ future request information within a limited future time window. In addition, it addresses a cost model and the resulting cost-minimization user-scheduling problem in multi-tier fog computing networks. The chapter first presents the system model. Under the fog-enabled network architecture, the chapter formulates the problem under consideration. The online fog-enabled multi-tier operations scheduling (FEMOS) algorithm is then proposed and corresponding performance analysis is conducted. The chapter further develops the PMOS algorithm based on the proposed FEMOS algorithm and predicted users’ information. Furthermore, it proposes a unified multi-tier cost model to motivate the fog access nodes for resources sharing, and develops the cost-oriented user scheduling algorithm to effectively solve the resulted cost-minimization user scheduling problem.