Home // International Journal On Advances in Networks and Services, volume 17, numbers 3 and 4, 2024 // View article

Enhancing Path Reliability in Contact Graph Routing via Improved Hop Time Estimations

Authors:
Ricardo Lent

Keywords: delay-tolerant networking; routing; reliability; performance evaluation; cognitive networking

Abstract:
Delay-Tolerant Networking(DTN) enables the forwarding of data bundles over space networks that experience extended link disruptions and path disconnections. Routing in such environments is challenging yet crucial for efficient end-to-end data delivery. Contact Graph Routing (CGR) is the standard routing method adopted for space DTNs. This study enhances CGR by exploring the potential inclusion of a Cognitive Element (CE) that leverages a data-driven approach. The CE is anticipated to use machine learning to estimate average single-hop bundle delivery times based on selected inputs. These estimates then replace the propagation delay that is used as the sole decision metric in CGR's shortest-path algorithm, improving the accuracy of the average one-hop delivery time predictions by allowing consideration of significant factors such as Convergence Layer Adapter (CLA) behavior, configuration parameters, packet drop probabilities, and unreliable contacts. The result is enhanced routing performance. The paper evaluates the CE extension in a simulated Earth-Moon network, assuming implementation independence and examining the effects of contact plan size (defined as a look-ahead window) and potential performance degradation from reduced prediction accuracy due to partial data or model limitations. Insights into the practical benefits of this approach are provided with a focus on realistic contact features and unreliable links.

Pages: 95 to 104

Copyright: Copyright (c) to authors, 2024. Used with permission.

Publication date: December 30, 2024

Published in: journal

ISSN: 1942-2644