AI-Enabled Reliable Delay Sensitive Communication Mechanism in IoUT Using CoAP

The highly dynamic and harsh conditions in the underwater environment pose challenges in enabling reliable and delay-tolerant communication in the Internet of Underwater Things (IoUT). Moreover, the constrained IoUT nodes have limited energy and are not capable of handling multiple retransmissions in the case of packet losses. Therefore, a lightweight communication mechanism is required that can promise reliable communication without large delays. In this article, we propose an AI-enabled reliable delay-sensitive communication mechanism in IoUT using a constrained application protocol (CoAP). The proposed mechanism uses a reliable transmission mode of CoAP to enable reliable communication. In addition, to achieve delay-sensitive communication, we modify the default congestion control mechanism of CoAP. The proposed scheme transmits multiple copies of the same packet without binary exponential backoff (BEB). Furthermore, to reduce the overhead caused by the transmission of multiple copies, reinforcement learning (RL) is employed at the sink to learn the transmission behavior of each node. In this way, the optimal number of copies are found which are to be sent by each node considering the real-time environmental conditions. The simulation results show that the mean energy decay is less compared to the default CoAP. Also, the performance is significantly increased in terms of the average packet delivery ratio (PDR) and average data delivery delay.