Effectual Side Channel Attack Detection Using Convolutional Neural Network Optimized CMP-COA

Smart phones are more vulnerable to malicious assaults because of technological advancements. An attack known as a side channel attack (SCA) in which attacker exploits the physical implementation of a cryptosystem to obtain confidential information. There are several existing works focused on side channel attack but lacks an accuracy and timely detection through important features. To resolve this issue, this paper proposes the novel technique intends to detect the side channel attack that includes phases of pre-processing, feature extraction and classification. Initially, the input data is pre-processed under data normalization and the statistical and higher order statistical features are extracted and are subjected to SCA classification using Optimal Convolutional Neural Network (OCNN). To make the prediction more precise and accurate, the weight of Optimal Convolutional Neural Network are optimized with hybrid Coronavirus Mask Protection and Coati Optimization Algorithm to improve the accuracy in detection of SCA. The proposed technique is implemented in python and the performance metrics such as accuracy, error rate, and execution time are evaluated. The proposed model shows that it has accuracy of 96% and the lowest execution time and error rate of 2.5 sec and 0.04<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup> which is far better when compared to the existing methods.