Abstract
This paper proposes a secure and energy-efficient Internet of Things (IoT) model for e-health. The main objective is to secure the transmission and retrieval of biomedical images over IoT networks. To achieve this, the compressive sensing and five-dimensional hyper-chaotic map (FDHC) are utilized to encrypt the biomedical images. However, FDHC suffers from the hyper-parameter tuning problem. To resolve this, a \theta -non-dominated sorting genetic algorithm III ( \theta -NSGA-III) based FDHC is proposed. Initially, the initial attributes of FDHC are tuned using \theta -NSGA-III. Then, the tuned attributes are used by FDHC to generate the secret keys. The obtained secret keys are then utilized to permute and diffuse the input biomedical images. Permutation and diffusion operations are applied row-wise and column-wise on the input biomedical image to perform encryption. The proposed technique is sensitive towards the input images as the permutation and diffusion of each row and column depends upon the initial scrambled row and column. Experimental results reveal that the proposed technique outperforms the state-of-art image encryption techniques. The proposed framework encrypts and decrypts images at a higher speed, therefore, it can be used to secure the communication in green IoT networks.