RUDN University mathematicians with colleagues from Saudi Arabia and China proposed a model of an underwater sensor network with an unmanned vehicle. It allows monitoring, for example, the state of underwater gas pipelines. Compared to analogues, the new system consumes 8 times less energy and at the same time reduces signal delay time. The results are published in Drones.
An underwater wireless network of sensors is used to monitor underwater environments, search for leaks in gas pipelines and prevent natural disasters. They read the information and relay it along the chain to the final center. The operation of such a network consumes a lot of energy. In addition, the multiple data transmission between network nodes accumulates delay. An unmanned underwater vehicle that moves between sensors and collects information from them can solve these problems. However, for such a system to be effective, we need an optimal mode of operation and movement of the vehicle. RUDN mathematicians with colleagues from Saudi Arabia and China proposed such an algorithm.
“An underwater wireless sensor network is an environment with oceanographic sensors. These sensors are low-power devices that transmit received data using acoustic signals. Data is collected at a node on the shore or above water for further analysis. Such networks become more flexible when a moving drone is added to them . It either collects data from all sensors, or only from some head nodes. It is important to plan his route correctly,” said Ammar Muthanna, PhD, Director of the Research Center for Wireless 5G Networks Simulation of RUDN University.
In the proposed model, energy is saved in nodes through clustering. The nodes are combined into groups – clusters, with a schedule of work. The unmanned vehicle communicates between the clusters. The groups are formed by optimization algorithm – it selects the most convenient nodes that lead their groups. These nodes distribute the time that the sensors of the cluster spend in standby mode. The unmanned vehicle calculates its trajectory in such a way that it can simultaneously collect information from four clusters at the same time.
RUDN University mathematicians tested the new model in a computer simulation and compared the results with other solutions. Latency has been reduced to 40-80 milliseconds depending on the number of steps in the relay chain. For other models, this result is 50-120 milliseconds. Power consumption is practically independent of the number of sensors and remains at the level of 1 joule. Whereas for other models, power consumption can be 8 times higher and increases markedly with an increase in the number of sensors in the network.
“We plan to expand the created architecture by dividing the network. This will further reduce the power consumption of the sensors. In addition, we are going to test the architecture on a specific application for underwater wireless sensing systems,” said Ammar Muthanna, PhD, Director of the Research Center for Wireless 5G Networks Simulation of RUDN University.
