Acoustic channel impairments, long propagation delays, limited bandwidth, and high energy requirements all pose severe challenges to Underwater Internet of Things (IoUT) networks and thus, efficient routing is a critical design issue. State of art routing protocols are commonly based on a set of static heuristics and single-objective optimization techniques or learning protocols with heavy computation requirement and slow convergence that restrict their usability in dynamic underwater topographies. To overcome those constraints, this paper creates a bio-inspired adaptive optimization framework of energy-conscious routing in underwater IoT networks, in which routing is modeled as a multi-objective optimization problem, minimizing both energy usage and end-to-end delay and maximizing link reliability. The suggested framework combines biologically inspired evolutionary operations with adaptive parameter regulation to provide a dynamic exploration and exploitation balance when discovering a route and evolving. Also, unlike the routing scheme based on deep reinforcement learning, the proposed method converges fast and does not require a large amount of training data, so it can be applied to resource-limited underwater sensor nodes. Simulations show that the proposed framework conserves 18–27% of the average energy use, 14–22% of the end-to-end delay and 9% of the packet delivering ratio over recent metaheuristic- and learning-based routing protocols. Moreover, the network lifetime is increased by 20–30%, and routing overhead, and hop count are also reduced to a significant degree. Convergence analysis proves the ability to stabilize much faster and become more robust in dynamic acoustic conditions. These findings demonstrate that the suggested framework was an efficient and scalable routing solution to the next-generation underwater IoT applications.
Simon et al. (Sun,) studied this question.