Preventing Differential Cryptanalysis Attacks Using a KDM Function and the 32-Bit Output S-Boxes on AES Algorithm Found on the Internet of Things Devices

Mbuyu Sumbwanyambe,Khumbelo Difference Muthavhine

doi:10.3390/cryptography6010011

Mbuyu Sumbwanyambe, Khumbelo Difference Muthavhine

Open Access

https://doi.org/10.3390/cryptography6010011

Copy DOI

Journal: Cryptography	Publication Date: Feb 22, 2022
Citations: 1	License type: CC BY 4.0

Affiliation: University of South Africa

Abstract

Many Internet of Things (IoT) devices use an Advanced Encryption Standard (AES) algorithm to secure data stored and transmitted during the communication process. The AES algorithm often suffers DC (DC) attacks. Little has been done to prevent DC attacks, particularly on an AES algorithm. This study focuses on preventing Differential Cryptanalysis attacks. DC attacks are practiced on an AES algorithm that is found on IoT devices. The novel approach of using a Khumbelo Difference Muthavine (KDM) function and changing the 8 × 8 S-Boxes to be the 8 × 32 S-Boxes successfully prevents DC attacks on an AES algorithm. A KDM function is a newly mathematically developed function, coined and used purposely in this study. A KDM function was never produced, defined, or utilized before by any researcher except for in this study. A KDM function makes a new 32-Bit S-Box suitable for the new Modified AES algorithm and confuses the attacker since it comprises many mathematical modulo operators. Additionally, these mathematical modulo operators are irreversible. The study managed to prevent the DC attack of a minimum of 70% on AES and a maximum of 100% on a Simplified DES. The attack on the new Modified AES Algorithm is 0% since no S-Box is used as a building block.

Highlights

Internet of Things (IoT) devices and platforms are advancing boundless while initiating a seamless combination of computer networks with things or objects [1,2]
This study aims to solve the problem of a DC attack used in IoT devices by intruders to discover the cryptographic keys of an Advanced Encryption Standard (AES)
The results showed that the global constraint permitted the calculation of Maximum Differential Characteristics (MDCs) at a higher speed than advanced models for single-key and related-key cryptanalysis attacks on Midori, and for single-key cryptanalysis attacks on an AES

Summary

Introduction

IoT devices and platforms are advancing boundless while initiating a seamless combination of computer networks with things or objects [1,2]. IoT is an open network platform and a new communication standard for the latest innovations, connecting multiple heterogeneous devices to render new conventional services [3,4]. Security and privacy are the crucial predicaments for the IoT devices and yet admit some of the immense inconveniences such as DC attacks [3,4]. IoT devices and platforms, with no skepticism, depend on cryptographic algorithms such as AES for the security and privacy of confidential information and data [3,4]. New services provided by IoT devices have to be sufficiently secured utilizing solid cryptographic algorithms such as AES [1,2]. While an improvement of security and privacy on IoT devices is observed, there is an increasing use of old cryptographic algorithms such as AES. AES has been implemented on other IoT devices to secure data used for online transactions such as smart cards [1]

Objectives

Methods

Results

Conclusion