ECOWN: An Optimization Framework for Carbon-Aware Routing in Wired Networks

As the climate crisis intensifies, the ICT sector faces increasing pressure to reduce its environmental impact. Computer networks, which form the backbone of modern digital services, contribute significantly to global energy consumption and carbon emissions, especially when powered by fossil-based electricity in geographically diverse regions. While renewable energy adoption is growing, its availability remains uneven and insufficient to fully decarbonize large-scale infrastructures. Consequently, reducing the energy consumption of network devices, particularly in high-carbon-intensity areas, becomes a critical short-term objective. A key strategy is using Software Defined Networks (SDN), which enable better device management, to adapt routing and network management to consider both traffic demands and the carbon footprint associated with energy use. In this paper, we propose ECOWN (Efficient Carbon Optimization for Wired Networks), an ILP-based framework that minimizes the carbon emissions of wired networks. ECOWN jointly optimizes routing and link activation by leveraging realworld data on traffic patterns, carbon intensity, and router power models. Through simulations over the Abilene topology, ECOWN achieves up to 59 % of reduction in carbon emissions compared to traditional shortest-path routing schemes.