Underwater acoustic sensor networks (UWASNs) have been presented as an advanced technology for detection and extraction of data in marine environments. UWASNs have many current real-time applications not limited to: resource exploration, seismic monitoring, marine explorations, oil and gas inspection, and military surveillance applications. However, this advanced technology is constrained to data detection, transmission, and forwarding. In addition, transmitting and receiving large volumes of data is requires an exhaustive amount of time and substantial power to be executed, and still fails to meet real-time constraints. This has directed our research focus to the development of a real-time underwater computing system to meet the required real-time constraints. In our research activities, we discover and extract valuable information from beneath the ocean using data-mining approaches. Previously, we introduced a set of real-time underwater system architectures (RTUSAs) that can address various network configurations according to their applications and data size We also presented architectures that use efficient data-gathering and information-extraction approaches to meet the required real-time constraints for underwater big-data applications. In this study, we extend our results and develop multiple-gateway placement topologies using geometric distribution characteristics to satisfy real-time constraints. The system performance has minimal end-to-end delay, and decreased power consumption. Finally, the simulation results are verified to validate the performance of our system according to multiple-gateway topologies.
