A Low Power Cryptography Solution Based on Chaos Theory in Wireless Sensor Nodes

Abstract

Unobtrusive personal data collection by wearable sensors and ambient monitoring has increased concerns about user privacy. Applying cryptography solutions to resource constraint wireless sensors as one of the privacy-preserving solutions demand addressing limited memory and energy resources. In this paper, we set up testbed experiments to evaluate the existing cryptographic algorithms for sensors, such as Skipjack and RC5, which are less secure compared to block cipher based on chaotic (BCC) on existing IEEE802.15.4 based SunSPOT sensors. We have proposed modified BCC (MBCC) algorithm, which uses chaos theory characteristics to achieve higher resistance against statistical and differential attacks while maintaining resource consumption. Our comparison observations show that MBCC outperforms BCC in both energy consumption and RAM usage and that both MBCC and BCC outperform RC5 and Skipjack in terms of security measures, such as entropy and characters frequency. Our comparison analysis of MBCC vs BCC suggests 13.44% lower RAM usage for encryption and decryption as well as 6.4 and 6.6 times reduced consumed time and energy for encrypting 32-bit data, respectively. Further analysis is reported for increasing the length of MBCC key, periodical generation of master key on the base station and periodical generation of round key on the sensors to prevent the brute-force attacks. An overall comparison of cipher techniques with respect to energy, time, memory and security concludes the suitability of MBCC algorithm for resource constraint wireless sensors with security requirements.