An Autonomous Evolutionary Approach to Planning the IoT Services Placement in the Cloud-Fog-IoT Ecosystem

摘要

Abstract The growth of network-based computing devices and the development of massive services are both results of the Internet of Things (IoT) by deploying Internet-connected devices at the edge. These devices are limited in terms of storage and computing, where the resources they need are provided by the cloud computing paradigm. Unfortunately, the two-tier cloud-IoT architecture is not efficient enough to provide the resources needed for latency-sensitive applications. Consequently, Fog Computing Paradigm (FCP) has been proposed to complement cloud computing and support such IoT-generated applications at the network edge. Heterogeneity, geographical distribution, and large-scale of fog nodes need the development of new methodologies for deploying and running IoT applications on fog nodes. A collection of IoT services, which have varying Quality of Service (QoS) needs, make up applications for the IoT, so finding an autonomous IoT Fog Service Placement (IoT-FSP) scheme in such an infrastructure can be challenging. The current study presents a distributed conceptual computing framework to address this problem. It is based on an autonomous approach, which improves resource management in three-tier cloud-fog-IoT architecture. Besides, we use the Cuckoo Optimization Algorithm (COA) as a meta-heuristic approach to efficiently solve the FSP. We refer to the proposed strategy as the FSP-COA. The FSP-COA formulates the problem as a multi-objective problem to reconcile various objectives, including SLA violations, service latency, response time, fog utilization, service cost, and energy consumption. Finally, the performance of FSP-COA is evaluated compared to state-of-the-art algorithms using the iFogSim simulator. According to experiments, FSP-COA is more efficient than other algorithms regarding various metrics, including latency, energy and cost, and on average, outperforms the better of existing algorithms ranging from 4% to 16%.