Energy Measurement and Analysis of Security Algorithms for Embedded Systems

Energy and security are hot topics for real-time embedded systems. However, little traditional research consider the two factors together, which leaves a great challenge to guarantee the security of embedded applications running under battery-powered real-time embedded systems. In this paper, we investigate the characteristics of energy consumption for the most popular security algorithms. To precisely measure the energy cost of various cryptographic algorithms, we build a real measurement platform based on ARM9 and μC/OSII. We mainly test confidentiality and integrity algorithms included in a public cryptlib. Based on the experimental results, we conduct the energy features of asymmetric/symmetric algorithms, hash algorithms, HMAC algorithms and the security parameters. Furthermore, some recommendations about the utilization of cryptographic algorithms are proposed, which may be very valuable for running security applications in energy constrained embedded systems.