Performance Evaluation of Cryptographic Algorithms on ESP32 with Cryptographic Hardware Acceleration Feature

University essay from KTH/Skolan för elektroteknik och datavetenskap (EECS)

Abstract: The rise of the Internet of Things (IoT) and autonomous robots/vehicles comes with a lot of embedded electronic systems. Small printed circuit boards with microcomputers will be embedded almost everywhere. Therefore, the security and data protection of those systems will be a significant challenge to take into consideration for the future development of IoT devices. Cryptographic algorithms can be used to provide confidentiality and integrity for data transmitted between those embedded devices. It is important to know what kind of algorithm is the most suitable for the specified task and the selected embedded device.  In this thesis, several commonly used cryptographic algorithms are evaluated and an EPS32 based IoT device is chosen as the evaluation platform. ESP32 is a series of low cost and low power System-on-Chip microcontrollers with integrated Wi-Fi and dual-mode Bluetooth. Additionally, ESP32 has the hardware acceleration feature for commonly used cryptographic algorithms. The goal of this thesis is to evaluate the performances of different cryptographic algorithms on the ESP32 with and without using the hardware acceleration feature. The execution times of different cryptographic algorithms processing data with varying sizes are collected, and the performance of each cryptographic algorithm is then evaluated.  A data logging scenario is evaluated as a case study where the ESP32 periodically sends data to a remote database. Under different configurations of the ESP32, the transmission time of encrypted and non-encrypted communications via Hypertext Transfer Protocol Secure (HTTPS) and Hypertext Transfer Protocol (HTTP) will be compared.  The results can be used to simplify the calculation of performance/protection trade-offs for specific algorithms. It also shows that the built-in hardware acceleration has a significant impact on increasing those algorithms’ performances. For Advanced Encryption Standard (AES), the throughput for encryption increased by 257.8%, and for decryption 222.7%. For Secure Hash Algorithm (SHA-2), the throughput increased by 165.2%. For Rivest-Shamir-Adleman (RSA), the encryption throughput has a decrease of 40.7%, and decryption has an increase of 184%. Furthermore, the results can also aid the design and development of a secure IoT system incorporating devices built with ESP32. 

  AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)