A State-of-the-Art Survey and Taxonomy of Classification, Algorithms, and Techniques for Load Balancing in SDN

Abstract

The management of traditional networks has become increasingly complex due to the expansion of the network and the development of new technologies such as cloud computing, the Internet of Things (IoT), and big data. Therefore, it is imperative to transition from operating within conventional networks to utilizing advanced networks capable of effectively managing modern technology. One of the most significant advancements in networking is the implementation of software-defined networks (SDN). SDNs aim to decouple the control plane, which controls network functions, from the data plane, which handles data transmission. This separation enhances the flexibility of network management. The distribution of traffic inside SDN networks plays a crucial role in enhancing network performance and response. Implementing Load Balancing (LB) enhances overall system performance and guarantees the efficient and dependable utilization of network resources. This research aims to comprehensively analyze recent research studies and the taxonomy of LB in SDN, such as classification, algorithms, and techniques. This research provides a comprehensive, state-of-the-art survey of LB in SDN according to LB-Classification, LB algorithms, and LB-Techniques. This research proposed a modern taxonomy for LB-Classification based on two factors: scheduling and models. Also, it proposed a new taxonomy of LB-Algorithms based on three types (static, dynamic, and hybrid) and a taxonomy for a third type (hybrid) consisting of three kinds (hybrid-LB, hybrid dynamic-LB, and hybrid static-LB). Finally, this research proposed a modern classification of LB-Techniques based on six types: (Controller -LB, Server -LB, Path Selection and Re-route - LB, Scheduling Management and Queue -LB, Artificial Intelligence -LB, and Wireless and Wi-Fi-LB).