Machine learning based cryptanalysis techniques: perspectives, challenges and future directions

Abstract

The field of cryptanalysis has lately witnessed considerable advancement due to the need for artificial intelligence technologies to simplify the complicated task of vulnerability assessments for cryptographic algorithms. The use of well-known tools such as machine learning and deep learning has piqued the interest of researchers and experts in the field because it has supported research work in discovering great knowledge on the strong and weak points of cryptographic techniques while ushering in the era of automated and AI-driven cryptanalysis.Despite the positive solutions obtained through using DL in the realm of cryptanalysis, it is not without drawbacks. This paper emphasizes the issues encountered when using ML and DL in cryptanalysis as well as new paths of DL with the advent of the quantum neural network approach, which can provide better answers and hence the relevant state of the art.