Dynamic Optimization of Slot Management MAC Protocol for Large-Scale IoUT Based on POMDP

The development of the Internet of Underwater Things (IoUT) is of great significance to marine scientific research, deep-sea exploration and interdisciplinary data fusion. However, the existing medium access control (MAC) protocols usually face serious network congestion and unfair channel resource allocation challenges in large-scale IoUT, resulting in a decline in the overall performance of the system. To address these problems, this paper proposes a dynamic slot management MAC protocol based on POMDP (P-DSM-MAC). The protocol integrates network load estimation, ACB scheme optimization, dynamic slot allocation, and the sensing and multiplexing of idle/collision sub-slots through slot division to achieve efficient network resource management and sub-slot contention collision control. Simulation results show that P-DSM-MAC is significantly superior to existing protocols in key performance indicators such as network throughput, delay, and sub-slot contention collision rate, providing a feasible solution for the intelligent and dynamic optimization of IoUT in the future.