Intelligent Hyper-Heuristic Algorithm for Optimizing Application Placement in Fog-Healthcare

Abstract

Fog computing has become essential for solving latency-constrained applications, especially for instances of healthcare monitoring. However, efficient placement of the applications on the fog network still remains a problem due to factors such as resource constraints, as well as fluctuating workload. The traditional models have limitations in terms of the effective utilization of resources and minimizing latency. To overcome such limitations and to have a faster processing rate, this work presents a hyper-heuristic optimization-based algorithm for healthcare fog computing. For the development of an ECG (electrocardiogram) application, latency is considered a primary optimization target while meeting the resource constraints of the Fog nodes. The proposed hyper-heuristic algorithm can choose simple low-level heuristics based on the network conditions and workloads. Simulations are conducted, and the proposed model is compared against the default FCFS (First-Come, First-Served) policy and stateof-the-art algorithms in terms of delay, fog utilization, application admission rates, energy consumption, and total cost. The results clearly show that the proposed hyper-heuristic outperforms existing methods by up to 70% in terms of fog utilization and close to ~98% admission rates with significantly lowered delay. The outcomes show that using a hyper-heuristic for application placement is effective, especially for real-time healthcare applications such as ECG monitoring.