Sustainable and Secure Optimization of Load Distribution in Edge Computing

Abstract

Nowadays, the Internet of Things (IoT) plays a disruptive role in society by bringing benefits to different industries. IoT initiatives (e.g., transportation and manufacturing) are becoming more popular, and new applications are expected in the next few years. In this sense, Cloud Computing enables small IoT devices to exceed their limited processing power and, to complement this concept, Edge Computing introduces edge processing services with short communication delay, providing services and performing calculations closer to the network and data generation. This paradigm enables systems to operate under more restrictive requirements. However, load balancing is challenging due to the various devices in IoT environments that need to be connected to different edge servers. Besides, the heterogeneity of such servers poses another obstacle to be overcome to achieve efficient load balancing capabilities. The main goal of this research is to propose a sustainable and secure load balancing approach for edge computing using Particle Swarm Optimization (PSO). This proposal distributes IoT devices connections across multiple edge servers while observing sustainability and security KPIs. To accomplish this, we utilize PSO to assign IoT devices to edge servers aiming at balancing load, sustainability, and security KPIs. Besides, we compare the results of this proposal with different optimization methods in different experiments. The results showed that our proposal outperforms those methods in all cases investigated since it can observe different metrics in balancing the load of edge servers.