T‐RESIN: Throughput‐aware dynamic resource orchestration for Internet of Everything‐enabled software‐defined edge networks

Abstract

SummaryIn this work, we address the problem of resource allocation in Internet of Everything (IoE)‐enabled software‐defined edge networks. In the existing literature, the researchers considered optimizing the performance of the software‐defined networking (SDN) platform using a single‐tier architecture, where the Internet of Things (IoT) devices are in the same tier. However, with the advent of edge computing, we can explore the two‐tier architecture of edge networks—local tier and edge tier—in the presence of SDN, which has not been explored. Hence, we propose an evolutionary game‐based resource allocation scheme for software‐defined edge networks. Additionally, we aim to optimize the QoS of the edge‐based IoE services while optimizing the throughput of the system. The IoT devices use the proposed scheme in the local tier to identify the optimized mapping to the SDN switches. On the other hand, in the edge tier, the proposed scheme aims to optimize the throughput while allocating the IoE service to the optimal subset of edge devices. We evaluated the performance of the proposed scheme using the Python3‐based Mininet platform in the presence of Ryu Controller and Open vSwitches. Ryu is an element‐based software defined networking framework, which offers software elements with well‐described API that enables developers to produce new network control functions. We observed that by using T‐RESIN, the network throughput increases by 27.98%–31.84% than using the existing schemes.