New Privacy-Preserving Routing Method Cuts Multi-Domain Path Calculation Time by 24%, Keeps 90% Bandwidth Efficiency | Newswise

Researchers have developed a novel algorithm, BI-TE, that allows multi-domain networks to collaborate on traffic management without exposing sensitive internal data. By blending differential privacy with Graph Neural Networks, the system enables operators to share only perturbed network information. This approach effectively addresses the long-standing dilemma where providers must choose between maintaining high performance and protecting proprietary topology details. The method introduces controlled noise to local network reports, masking precise internal structures while allowing a central controller to determine efficient routing paths. A machine learning model compensates for this privacy-induced inaccuracy by predicting actual bandwidth usage from the noisy inputs. This mechanism ensures that cross-domain cooperation remains secure, as individual domains never reveal their private network configurations or resource specifics to external entities. This advancement is significant for open data and network transparency because it proves that robust privacy safeguards and operational efficiency are not mutually exclusive. It offers a practical framework for ISPs and enterprises to optimize traffic engineering and meet service guarantees collaboratively. By demonstrating that accurate, high-speed routing can coexist with strict data confidentiality, BI-TE sets a new standard for secure interoperability in complex, multi-stakeholder digital infrastructures.

Source: newswise.com
Published on 2025-07-24

New Privacy-Preserving Routing Method Cuts Multi-Domain Path Calculation Time by 24%, Keeps 90% Bandwidth Efficiency | Newswise

Related news