Cross-Medium Throughput Optimization With AUV Mobile Relay Path Planning

Reliable underwater-to-onshore information transmission is becoming essential for ocean ecological monitoring, marine geological survey, and underwater infrastructure inspection. Traditional static buoy relays lack the flexibility to adapt to spatially varying underwater acoustic (UWA) channels and mitigate UWA/radio frequency (RF) link rate mismatch, creating a bottleneck for cross-medium throughput enhancement. To address this problem, we propose a cross-medium throughput optimization scheme via AUV mobile relay path planning. We maximize the end-to-end throughput of the dual-hop cross-medium network by designing an optimal relay path in a quantified spatially varying UWA channel. We then develop two novel path planning algorithms based on ant colony optimization (ACO) and reinforcement learning (RL), with flexible destination settings for diverse underwater environments. The simulation results demonstrate that the proposed schemes outperform conventional relay systems and achieve superior optimization in end-to-end cross-medium throughput. Our proposed algorithms also exhibit stronger robustness to channel spatial variations and slighter performance degradation at increasing air receiver distances. This work enables cross-medium performance enhancement for sea-to-air information delivery.