LEEV: Load-Aware and Energy-Efficient VNF Deployment in NFV-Enabled Networks

Abstract

Network function virtualization (NFV) decouples network functions from dedicated hardware by implementing them as software-based virtual network functions (VNFs), thereby enabling flexible network resource management. Network services are provisioned through service function chains (SFCs) composed of multiple VNFs. Given SFC requests and physical machines (PMs), this paper investigates the VNF deployment problem with three objectives: maximizing the service acceptance ratio (SAR), minimizing the total energy consumption (TEC), and balancing the load among PMs. We then propose a load-aware and energyefficient VNF deployment (LEEV) scheme that dynamically adapts its deployment strategy based on the system load. Under lightload conditions, LEEV deploys adjacent VNFs of a given SFC on one PM to reduce energy consumption, while under heavy-load conditions, it can shift toward load-balancing deployment without sacrificing energy efficiency. Simulation results reveal that LEEV achieves a higher SAR, lower TEC, and improved load balance.