Multihop Age-of-Information-Enhanced Routing Mechanism for Delay-Sensitive Service in the LEO Satellite Networks

Low Earth Orbit (LEO) satellite networks, as a crucial component of sixth-generation (6G) mobile communication systems, offer seamless global coverage and ubiquitous connectivity. However, their large-scale constellations and high dynamic environments present significant challenges for conventional static and quasi-static routing protocols, particularly in meeting the stringent information freshness requirements of time-sensitive services such as disaster early warning. To address these challenges, we propose an adaptive multipath routing strategy aimed at minimizing the age of information (AMRS-MAoI). First, by leveraging a feedback-based retransmission mechanism with constrained retransmission attempts, we formulate the routing optimization as an average age of information (AoI) minimization problem. To improve resilience amid network congestion and intermittent link disruptions, we introduce a disruption-tolerant multipath routing algorithm that proactively discovers multiple disjoint backup paths to ensure uninterrupted real-time data updates. This approach not only reduces AoI but also enhances link utilization efficiency. Building on this, we design a deep reinforcement learning (DRL)-based adaptive routing framework with traffic segmentation, which dynamically optimizes data generation rates and traffic segmentation ratios across available paths to reduce packet loss and further improve information freshness. Extensive simulations demonstrate that, relative to existing techniques, our method substantially lowers packet loss rates and increases throughput, achieving an average AoI reduction of 11.25%, thereby significantly enhancing the timeliness and reliability of service data delivery.