With the advent of 5G, network slicing brings flexibility for heterogeneous services but also increases concerns about the energy demand of virtualized infrastructures. In France, the Haut Conseil pour le Climat estimates that CO2 emissions from communication networks may rise by 16--42% between 2020 and 2030, representing 16 to 40~TWh of annual energy consumption.

This work examines how to balance service performance and energy efficiency when allocating slices under strict SLA constraints. A preliminary review highlights that, despite extensive research on energy-aware NFV, the impact of routing decisions on total energy consumption remains insufficiently explored when combined with VNF placement.

To address this gap, we implement A*-based heuristic methods inspired by prior work on VNF placement and extend them to jointly determine placement and routing under a load-dependent energy model. In parallel, a Mixed Integer Linear Program is being developed, although numerical results are not yet available.

This study aims to clarify the trade-offs between performance and energy consumption in 5G slice orchestration and contribute to the design of more energy-efficient next-generation networks.