A Unified Cross-entropy Based Task Scheduling Algorithm for Heterogeneous Fog Networks
Zening Liu, Yang Yang, Ming‐Tuo Zhou, Ziqin Li · 2018
Fog computing, which shifts functions of clouds to the network edge along the cloud-to-things continuum, has risen as a promising architecture for realizing the vision of smart city. In fog computing, how to efficiently map computation tasks into a variety of heterogeneous fog nodes (FNs) to improve the quality of service, such as low latency and energy consumption, is a fundamental challenge. In this paper, we investigate this task scheduling problem in general multiuser multi-FN hetero- geneous fog networks. We first show that such a problem can be formulated as a typical combinational optimization problem with different objectives. Then, we study this problem utilizing the cross-entropy (CE) method through associating it with a rare event estimation problem, and propose a unified and efficient CE based task scheduling algorithm. Simulation results demonstrate that the proposed CE based algorithm can achieve near-optimal performance in a fast manner, which much outperforms other baseline solutions