This thesis explores the feasibility of embedding carbon-intensity metrics into inter-domain routing to reduce end-to-end emissions while maintaining BGP correctness and performance. By integrating CAIDA’s PoP-level Internet topology with Electricity Maps’ carbon data, the study evaluates a Lowest Emissions (LE) strategy against traditional Shortest Path (SP) routing under valley-free constraints. The research begins by motivating carbon-aware routing through a comprehensive literature review that identifies current gaps in sustainable networking. The methodology chapter details the geolocation of Points-of-Presence, their mapping to regional power grids, and the enforcement of path validity, followed by a pilot test on a European subnet. The core experimental analysis provides a global-scale comparison between SP and LE, quantifying the trade-off between emission reduction and path stretch. Results demonstrate that carbon-aware selection achieves consistent savings, though gains are heavily influenced by topological flexibility and grid heterogeneity. The study further investigates alternative scenarios, such as zero-emission critical ASes and the relationship between customer cone size and carbon intensity. Finally, the work addresses the BGP threat surface introduced by carbon signaling and proposes cryptographic frameworks to secure these metrics. The thesis concludes by summarizing key trade-offs and outlining future research directions for the operational integration of carbon-aware signals.
This thesis explores the feasibility of embedding carbon-intensity metrics into inter-domain routing to reduce end-to-end emissions while maintaining BGP correctness and performance. By integrating CAIDA’s PoP-level Internet topology with Electricity Maps’ carbon data, the study evaluates a Lowest Emissions (LE) strategy against traditional Shortest Path (SP) routing under valley-free constraints. The research begins by motivating carbon-aware routing through a comprehensive literature review that identifies current gaps in sustainable networking. The methodology chapter details the geolocation of Points-of-Presence, their mapping to regional power grids, and the enforcement of path validity, followed by a pilot test on a European subnet. The core experimental analysis provides a global-scale comparison between SP and LE, quantifying the trade-off between emission reduction and path stretch. Results demonstrate that carbon-aware selection achieves consistent savings, though gains are heavily influenced by topological flexibility and grid heterogeneity. The study further investigates alternative scenarios, such as zero-emission critical ASes and the relationship between customer cone size and carbon intensity. Finally, the work addresses the BGP threat surface introduced by carbon signaling and proposes cryptographic frameworks to secure these metrics. The thesis concludes by summarizing key trade-offs and outlining future research directions for the operational integration of carbon-aware signals.
Carbon-Aware Routing in the Wild: Performance and BGP Security Considerations
LAINI, ANNALISA
2024/2025
Abstract
This thesis explores the feasibility of embedding carbon-intensity metrics into inter-domain routing to reduce end-to-end emissions while maintaining BGP correctness and performance. By integrating CAIDA’s PoP-level Internet topology with Electricity Maps’ carbon data, the study evaluates a Lowest Emissions (LE) strategy against traditional Shortest Path (SP) routing under valley-free constraints. The research begins by motivating carbon-aware routing through a comprehensive literature review that identifies current gaps in sustainable networking. The methodology chapter details the geolocation of Points-of-Presence, their mapping to regional power grids, and the enforcement of path validity, followed by a pilot test on a European subnet. The core experimental analysis provides a global-scale comparison between SP and LE, quantifying the trade-off between emission reduction and path stretch. Results demonstrate that carbon-aware selection achieves consistent savings, though gains are heavily influenced by topological flexibility and grid heterogeneity. The study further investigates alternative scenarios, such as zero-emission critical ASes and the relationship between customer cone size and carbon intensity. Finally, the work addresses the BGP threat surface introduced by carbon signaling and proposes cryptographic frameworks to secure these metrics. The thesis concludes by summarizing key trade-offs and outlining future research directions for the operational integration of carbon-aware signals.| File | Dimensione | Formato | |
|---|---|---|---|
|
Tesi878279.pdf
accesso aperto
Dimensione
9.51 MB
Formato
Adobe PDF
|
9.51 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14247/28248