Dynamic Resource Orchestration for Edge Computing in SDN Environments

Edge computing has emerged as a critical paradigm for supporting latency-sensitive applications in Software-Defined Networking (SDN) environments. This paper proposes a dynamic resource orchestration framework that optimizes task offloading and resource allocation at the edge using a distributed optimization approach. By integrating Lyapunov optimization with SDNs centralized control, our method minimizes latency and energy consumption while ensuring quality of service (QoS). We formulate the problem as a Mixed Integer Non-Linear Programming (MINLP) model and propose a heuristic-based relaxation to enhance scalability. Simulations on edge-cloud topologies demonstrate that our approach achieves lower latency and higher resource efficiency compared to centralized methods, validated through extensive performance evaluations.