Dual Security based on Crypto Steganography Using Two Level Learning In Assist with Dual Offbeat Shielding Design

Abstract

A sensor node and other electronic devices attached to any IoT object can be involved in the communication over wireless network in IoT environments, which makes it necessary to preprocess a large amount of sensed data before storing it. Therefore, sensing data in the form of images is to be sent to the cloud storage system via wireless medium, but this suffers from image hijacking where data is manipulated, which leads to insecure transmission. To mitigate this problem, two levels of security are employed. Memory retaining is the primary level of enhancing learning which uses past experiences to extract optimal features from sensed images and then subjected to offbeat shielding activities, which include cryptographic steganography. The proposed system creates cloud storage that is protected by the optimal features learned from a neural network, thus ensuring that clouds are secure in the Internet of Things.