Resource orchestration in software-defined edge networks for IOT

Abstract

In recent years, the proliferation of Internet of Things (IoT) applications has resulted in a substantial increase in heterogeneous and dynamic data traffic, demanding stringent quality of service (QoS) guarantees such as high throughput, low delay, strong reliability, and energy efficiency. Traditional network architectures with static configurations and vendor-specific constraints are not capable of handling dynamic traffic patterns. Software-Defined Networking (SDN) integrated with edge computing offers a programmable and flexible networking paradigm with centralized control. However, achieving effective resource orchestration across Software-Defined Edge Networks (SDEN) and softwarized 5G/6G environments remains challenging due to heterogeneous traffic characteristics, dynamic IoT workloads, and limited network resources. We propose a comprehensive set of analytical and heuristic orchestration frameworks to address throughput optimization, delay reduction, traffic management, and energy efficiency in SDEN and softwarized 5G/6G networks. The first work introduces T-RESIN, a throughput-aware resource orchestration framework that uses an evolutionary game to determine equilibrium-driven flow distribution across SDN switches and edge nodes. Building on this, D-RESIN presents a delay-aware orchestration mechanism based on an evolutionary game-theoretic approach to minimize processing delay for delay-sensitive IoT applications. For next-generation wireless networks, TRON is proposed as an SDN-based traffic management architecture that leverages OpenFlow Group Tables to dynamically balance heterogeneous traffic and improve link utilization.