A Secure Structure for Hiding Information in a Cryptosystem based on Machine-learning Techniques and Content-based Optimization using Portfolio Selection Data

Abstract

Many systems including networks environment have higher complexity and ubiquitous connections than a normal system, hence design of a security system for hiding pertinent data a challenging task. This paper presents a secure structure that extends a protocol which is available for fast Diffe-Hellman protocol using Kummer surface. We show the extended version of the scheme by inclusion of an additional point, such that a more secure system can be constructed. The scheme has been build by employing machine-learning technique to select an appropriate class from multiple set of surfaces. A brief discussion on inclusion of multiple surfaces and making a selection of a specific surface using NSGA II algorithm is also provided. In this paper, we provide a brief overview of AES-128 (AES also known as Rijndael). In the starting, a short overview of the AES is given. This paper also has a description for altering the key generation module in AES based upon a newly designed content-based matrix which is built from portfolio selection data. The matrix is constructed using some predefined factors  modifies the existing index which is computed based upon the context of the message. An optimization algorithm is employed for selecting specified entries from content matrix. These selected entries are used for altering the key generation algorithm in AES. The modified output obtained after altering the key generation scheme is provided in the paper. Lastly, a brief overview of LIM index, which is used as an index in cryptanalysis, is given. This paper has a description of the scheme to construct a more secure system that is capable of hiding the information with above-mentioned techniques.