Coded Caching Enabled D2D Content Delivery in UAV-Assisted Marine Edge Networks

Due to their flexible mobility and rapid deployment, Unmanned Aerial Vehicle (UAV)-assisted marine edge networks can effectively mitigate the limitations of conventional shore-based networks, such as limited coverage and high content access latency. Coded caching, which partitions content into smaller encoded segments distributed across multiple Autonomous Underwater Vehicles (AUVs) and enables Device-to-Device (D2D) transmission, further enhances delivery efficiency by alleviating backhaul dependence. Motivated by these advantages, this paper investigates coded caching-enabled D2D content delivery in UAV-assisted marine networks, aiming to jointly optimize UAV trajectories, caching decisions, and content access strategies to minimize request latency under long-term UAV energy constraints. Distinct from terrestrial or single-domain networks, the dual-hop acoustic-RF architecture, intermittent underwater connectivity, and scarce marine resources impose additional challenges for maintaining content availability and reliable D2D delivery. To address this NP-hard problem, we propose a novel Online Joint Coded Caching and Content Delivery (OJC3D) algorithm. Leveraging the Lyapunov optimization framework, the original long-term problem is decomposed into a per-slot real-time optimization problem, which is then solved through a three-stage approach that sequentially optimizes UAV trajectories, content caching, and content request decisions based on convex optimization theory. Simulation results demonstrate that, compared with benchmark schemes, the proposed OJC3D algorithm reduces content access latency by up to 20% and UAV energy consumption by 35%, achieving energy-efficient and low-latency content delivery in dynamic marine environments.