Q-learning Based Delay Sensitive Routing Protocol for Maritime Search and Rescue Networks

In recent years, maritime accidents increase with the frequent maritime activities. Real-time transmission of search and rescue video can help decision makers make better search and rescue arrangements. To improve the efficiency of the maritime search and rescue tasks, we propose a Q-learning based delay sensitive routing protocol (QDSR) for maritime search and rescue networks. QDSR can dynamically adapt to the changes of the topology of maritime communication networks and optimize the end-to-end delay and packet delivery ratio through iterative learning. Simulation results demonstrate that the proposed QDSR can select an optimal path, thereby improve the reliability and timeliness of maritime emergency video transmission.