Experimental performance analysis of lightweight block ciphers and message authentication codes for wireless sensor networks

• Main Authors: Sooyeon Shin, Minwoo Kim, Taekyoung Kwon
• Format: Article
• Language: English
• Published: SAGE Publishing 2017-11-01
• Series: International Journal of Distributed Sensor Networks
• Online Access: https://doi.org/10.1177/1550147717744169

Wireless sensor networks are widely used in many mission critical applications such as environment monitoring, military monitoring, and healthcare. The highly sensitive nature of collected information and various potential security threats make security in wireless sensor networks a crucial concern. For resource-constrained nature of these special networks, the design of lightweight cryptography has become active research topic over the last several years. In this article, we provide experimental performance analysis of several lightweight block ciphers and message authentication codes to help for choosing better algorithms for wireless sensor networks in terms of efficiency. Unlike the previous simulation-based benchmarking projects, we employ representative three real-world devices of wireless sensor networks: 8-bit AVR (Arduino Uno), 16-bit MSP (Tmote), and 32-bit ARM (Raspberry Pi 2) microcontrollers. We also introduce an efficient implementation of rotations that can improve the performance of lightweight block ciphers with 32-bit rotation operations on both 8-bit AVR and 16-bit MSP microcontrollers and verify this methodology. We evaluate and compare performance of selected block ciphers and message authentication codes to recommend better solutions for wireless sensor networks. Based on our results, we show that the simulation-based results could be inaccurate and different from results of real implementations. As a result of our experimental performance analysis, we identify lightweight block ciphers and message authentication codes suitable for wireless sensor networks and provide their feasibilities.