Artificial Intelligence-Enabled Fully Echoed Q-Routing and Adaptive Directional Medium Access Control Protocol for Flying Ad-Hoc Networks

In today's world, communication is critical to multi-UAV (Unmanned Aerial Vehicle) system design, enabling UAVs to collaborate and operate cohesively. UAVs generally rely on infrastructure-based communication through ground stations or satellites. However, this approach has numerous limitations, particularly in multi-UAV systems. Ad hoc networking among UAVs offers a solution by allowing direct communication without needing fixed infrastructure. This work introduces two innovative protocols: Artificial Intelligence-enabled Fully Echoed Q-Routing (AI-FEQ) and Position-Prediction-based directional MAC (PPMAC) protocols for improving the performance of multi-UAV systems. These protocols leverage AI techniques like unsupervised, supervised, and reinforcement learning to make intelligent decisions regarding topology formation, maintenance, and routing management. Furthermore, the proposed AI algorithm will enhance the development and sustainability of Flying Ad Hoc Network (FANET) topologies that help to enlarge the communication efficiency and reliability of multi-UAV systems. The simulation results reveal that the proposed AI-FEQ protocol achieves an impressive network density association of 90% and minimal data transmission latency of 4.9 s as compared to the existing protocols.