Análise e Otimização do Algoritmo Criptográfico AES para Utilização em Aplicações IoT

Abstract

The Internet of Things (IoT) universe is expanding every year, featuring applications in various sectors. However, this growth generates a massive amount of data that flows through the Internet daily, therefore, the concern for the security of this information is of paramount importance. Cryptography is a good strategy for data protection, however, classical cryptographic algorithms are highly costly in terms of performance and energy consumption, which is particularly detrimental for IoT applications, as most of them are implemented on devices with limited computational resources. This dissertation proposes optimizations to the AES cryptographic algorithm, specifically in the SubBytes and MixColumns stages, with the aim of enabling secure IoT applications on resource constrained devices, making it a lightweight algorithm. The proposed optimizations yield significant results in terms of performance, without compromising the security of the algorithm, as well as in terms of energy consumption. In terms of performance, the proposed version achieved an average reduction of 86.71% in execution time and, furthermore, reductions of 31.82% and 89.04% in program memory and dynamic memory allocation, respectively, compared to the original AES. Regarding security, a 50.41% avalanche effect was achieved, along with a better distribution of 0’s and 1’s in the ciphertext compared to the original version. Moreover, it received approval in the NIST tests. Finally, in terms of energy consumption, the optimized version proposed by this dissertation achieved reductions of 63.87% and 70.86% when executed on ESP-WROOOM-32 and Raspberry Pi Pico platforms, respectively, making it much more attractive for IoT applications.