NeuraLunaDTNet: Feedforward Neural Network-Based Routing Protocol for Delay-Tolerant Lunar Communication Networks
This work addresses routing challenges for lunar communication networks, but it appears incremental as it builds upon the existing PRoPHET protocol.
The paper tackles the problem of inefficient routing in delay-tolerant lunar communication networks by proposing NeuraLunaDTNet, a feedforward neural network-based protocol that enhances the PRoPHET routing protocol, resulting in improved efficiency in dynamically changing spatio-temporal graphs.
Space Communication poses challenges such as severe delays, hard-to-predict routes and communication disruptions. The Delay Tolerant Network architecture, having been specifically designed keeping such scenarios in mind, is suitable to address some challenges. The traditional DTN routing protocols fall short of delivering optimal performance, due to the inherent complexities of space communication. Researchers have aimed at using recent advancements in AI to mitigate some routing challenges [9]. We propose utilising a feedforward neural network to develop a novel protocol NeuraLunaDTNet, which enhances the efficiency of the PRoPHET routing protocol for lunar communication, by learning contact plans in dynamically changing spatio-temporal graph.