Remora Jaya Optimization-Enabled Deep Quantum Neural Network for Underwater Target Tracking Using Radar Images

Abstract

The Ocean boundaries can be protected by pursuing an underwater uncooperative target and permits for the exploitation of ocean resources. This research design a novel technique, named Remora Jaya Optimization (RJO)-enabled Deep Quantum Neural Network (DQN) by considering the impacts of an unknown underwater environment for effective underwater target tracking in radar images. Here, the image is modified effectively using the RJO algorithm because the input radar signal is applied to the image reconstruction process. The modified image is fed up to the gridding phase, where the image is partitioned into a number of grids. The feature extraction process is carried out to extract the significant features after the gridding step is over. This data augmentation method is carried out to increase the data dimensionality. Accordingly, the augmented result is forwarded to the DQN for target tracking, where the network is tuned efficiently by the algorithm of RJO, which is devised by the integration of Jaya Optimization and Remora optimization algorithm (ROA). Moreover, the RJO-based DQN has achieved a minimum Means Square Error (MSE) of 0.168 and maximum detection rate of 0.914, and minimum MSE of 0.168 based on Vertical-Vertical (VV) polarity. The proposed method showed higher effectiveness in detecting the underwater target system in the marine environment.