Q-Learning-based Energy-Efficient Custom Cooperative Routing Protocol for Underwater Wireless Sensor Network

Underwater Wireless Sensor Networks (UWSNs) are currently a pivotal focus in academic and industrial domains due to their diverse applications, such as disaster prevention, military security, environmental monitoring, data collection, scientific research, and industrial usage. The underwater area is very dense, and hence, exploring such a denser environment is difficult in the first place. To make this exploration easy, underwater sensors are used that can collect information from underwater and forward it to the base station where these data can be used for various purposes. The problem with UWSNs is that they have a very limited amount of energy, so optimizing the energy usage of the sensors will be beneficial. To deal with this, this paper proposed a Q-Learning-based Energy-Efficient Custom Cooperative Routing (QEECCR) protocol that uses a Q-learning technique to optimize the routing based on the energy levels of the sensors. The algorithm selects a node based on the Q-value of the node for forwarding data to the base station. The proposed routing protocol is compared with the QCMR routing protocol, and results showed that it consumes less energy compared to the QCMR. With an increasing number of nodes under the water, designing a manual routing for low energy consumption becomes hard. This proposed protocol can remove human intervention and can find the routing path with less time and with higher accuracy.