LPDC: Mobility-and Deadline-Aware Task Scheduling in Tiered IoT
Jianhua Fan, Jianwei Liu, Jin Chen, Jing Yang · 2018
Fog computing shifts data processing to network edge to improve end users' Quality of Experience (QoE) through task offloading. In a tiered IoT infrastructure, efficient optimization algorithms are needed to jointly schedule tasks on cloud, cloudlet, and IoT devices, in order to meet application deadlines and account for the mobility of end devices/users. However, these two problems are traditionally considered separately, yet their degrees of freedoms are clearly coupled. In this paper, we tackle this combinatorial problem of joint optimization and present a novel mobility-and deadline-aware task scheduling algorithm that leverages location prediction based on Dynamic Pattern Tree. The scheduling problem is shown to be a multi-dimensional 0-1 knapsack problem and is solved using an improved Ant Colony Optimization (ACO) with links' priorities. Simulations with real-world mobility trace show that the performance of our algorithm outperforms other existing scheduling algorithms.