Abstract

Software-Defined Networking (SDN) offers flexibility and programmability in network management, but efficient resource allocation remains a challenge due to dynamic traffic patterns and diverse service requirements. This paper proposes an Enhanced Ensemble Method (EEM) for optimizing resource allocation in SDN environments. EEM integrates multiple ensemble learning techniques, leveraging their complementary strengths to enhance prediction accuracy and robustness. The key contribution lies in the novel integration of ensemble methods tailored for SDN resource allocation, offering improved adaptability to changing network conditions and service demands. Evaluation on real-world SDN datasets demonstrates that EEM outperforms existing methods in terms of both resource utilization efficiency and service quality. Notably, EEM achieves significant improvements in network throughput, latency reduction, and resource utilization balance.