Security and Performance in IoT: A Balancing Act

• Main Authors: Luke E. Kane, Jiaming James Chen, Rebecca Thomas, Vicky Liu, Matthew Mckague
• Format: Article
• Language: English
• Published: IEEE 2020-01-01
• Series: IEEE Access
• Subjects: AES; Acorn; ChaCha; ciphers; cryptography; the~Internet of Things (IoT)
• Online Access: https://ieeexplore.ieee.org/document/9133521/

With predictions suggesting there will be 18 billion Internet of Things (IoT) devices live by 2022, performance of these low powered devices, as well as security is of utmost importance. Managing security and performance is a balancing act. Achieving this balance will always continue to be a challenge. This research presents two main contributions to this area. The first contribution is a framework to measure cryptographic performance of IoT devices. The areas of measurement are power consumption, time cost, energy cost, random access memory (RAM) usage and flash usage. The second contribution is an insightful comparison of the performance of the ATmega328, STM32F103C8T6 and ESP8266 low powered microcontroller devices. Experiments were conducted on these devices running various cryptographic operations. The measured operations are from three encryption algorithms: Advanced Encryption Standard (AES), ChaCha and Acorn. The proposed methods from this research are real-world in nature rather than simulated, and can be used by others wishing to conduct their own IoT performance testing. The results show that the ATmega328 has the lowest overall power consumption. The ESP8266 was generally the fastest performing device. ChaCha outperformed AES in both time cost and energy cost. Both algorithms outperformed Acorn in these metrics. The STM32F103C8T6 device displayed the best overall energy cost, while still performing well in terms of time. The results from the experiments conducted in this study can be used by network designers, developers and others to make appropriate decisions in IoT deployments with regards to balancing performance and security.