Dr. Shalabh Gupta, an electrical engineer at the University of Connecticut, is in the process of developing a more efficient and resilient sensor network that could be used by the US Navy to detect and track underwater targets of interests, such as enemy ships and submarines. Gupta’s approach is centered around a distributed algorithm that he had developed with his PhD student James Hare that would help isolate specific sensors that are within range of the targets of interest, and switch all other sensors nodes that are not within range to low power sensing or sleeping modes.
Dr. Gupta, an engineer and a researcher at the National Institute for Undersea Vehicle Technology, is working towards an energy minimization problem that is associated with the current sensor node technology that is in use. The navy employs thousands of individual underwater acoustic sensor nodes that are the building blocks of an underwater sensor network, that is used by the navy to track underwater target movement and oceanographic conditions. In the current technology, the underwater acoustic nodes are always on and tracking, even when there are no targets in site, which expends a large amount of energy. Gupta’s algorithm is focused around turning off those sensor nodes on and off opportunistically. When a target of interest comes within range, the sensor node that is closest to that target will switch on automatically and begin tracking the target at a high-power sensing mode. This sensor that has been switched on will also use the incoming information to predict targets’ trajectory and then alert other sensors that are in the projected trajectory of the incoming target. This will in turn, switch all other sensors in the target’s path to switch on to a high-power mode, while inhibiting all the other sensors not in the trajectory from turning on. These “off” sensors will be operating at a low-power or a sleep mode while they are not in use, which will greatly reduce the amount of energy used.
Since current sensor technologies run on full power, without turning on or off, the battery life for each sensor tends to be very short and given the location of the sensors, the process of replacing the batteries continuously due to a short life span is very time consuming and difficult. When a sensor goes out, there is a loss in coverage in that part of the ocean and targets can then go undetected. Gupta is looking to implement a system that will classify the target based on its movement and then inform the other sensors in the network. Movements would be analyzed in terms of their speed, direction and magnitude, and then would be classified as either a target of interest or not. This can also be used to help reduce the amount of energy used, as the network would only track when there is a target of interest in the area, and would not turn on if there was not one present.
While Gupta is still looking for funding to begin the construction of the underwater sensors, he has completed initial ground tests of sensor prototypes and has been in discussion with Navy officials about possibly implementing this promising technology into their underwater sensor networks.
