Rethinking Sustainable Campus Networking: Evidence from MST-Based Cabling Optimization and Device-Level Energy Consumption Measurements

Abstract

The digitalization of university campuses has intensified the need for reliable, energy-efficient, and sustainable network infrastructures. While Wi-Fi systems offer flexibility and ease of deployment, their long-term energy footprint often exceeds that of wired alternatives. Conversely, Ethernet networks, though more stable and efficient, require careful optimization to reduce installation costs and material use. This study develops a two-stage framework that combines graph-theoretic optimization with empirical device-level measurements to inform sustainable campus network design.In the first phase, the physical network layout was modeled as a Minimum Spanning Tree (MST) using the Manhattan distance metric to reflect realistic in-building cable routing. The MST was solved via Kruskal’s algorithm and validated through a mixed-integer programming model implemented in Gurobi, yielding a total cable length of 1,435.69 metres and a 63.4% reduction in total cable length—implying comparable savings in material and installation costs—compared with a naive layout. In the second phase, an experimental analysis examined the energy consumption of Wi-Fi, native Ethernet, and Ethernet-adapter connections across multiple devices. Although the paired t-tests revealed no statistically significant differences at the 95% confidence level, the results indicate that device-specific characteristics play a crucial role in determining actual energy efficiency.Overall, the findings underscore the importance of integrating optimization and empirical evaluation in sustainable network planning. The proposed framework not only supports cost-efficient infrastructure decisions but also contributes to the broader goal of reducing the carbon footprint of digital campuses.