Abstract
Encryption algorithms can be used to help secure wireless communications, but securing data also consumes resources. The goal of this research is to provide users or system developers of personal digital assistants and applications with the associated time and energy costs of using specific encryption algorithms. Four block ciphers (RC2, Blowfish, XTEA, and AES) were considered. The experiments included encryption and decryption tasks with different cipher and file size combinations. The resource impact of the block ciphers were evaluated using the latency, throughput, energylatency product, and throughput/energy ratio metrics. We found that RC2 encrypts faster and uses less energy than XTEA, followed by AES. The Blowfish cipher is a fast encryption algorithm, but the size of the plaintext affects its encryption speed and energy consumption. Faster algorithms seem to be more energy efficient because of differences in speed rather than differences in power consumption levels while encrypting.
1. Introduction
A personal digital assistant (PDA) is a handheld device that was originally designed as a personal organizer, but that has evolved into a wireless communication device. Many commercial and military applications now require PDAs with secure wireless transmission capability. Moreover, information privacy has become a major concern for all users, even personal users. A variety of algorithms are available for a user to encrypt data and, thus, prevent eavesdroppers from obtaining critical or private information. However, the execution of encryption algorithms consumes both time and energy. While certain encryption algorithms may be less vulnerable to compromise than others, constantly using cryptographically strong algorithms may result in severely reduced lifetimes for battery-powered devices such as PDAs. In other words, utilizing stronger encryption algorithms may consume more energy and drain the PDA battery faster than using less secure algorithms. Due to the processing requirements and the limited computing power in many PDAs, using strong cryptographic algorithms may also significantly increase the delay between data transmissions. Thus, users and, perhaps more importantly, software and system designers need to be aware of the benefits and costs of using various encryption algorithms. This research answers questions regarding energy consumption and execution time for various encryption algorithms executing on a PDA platform with the goal of helping software and system developers design more effective applications and systems and of allowing end users to better utilize the capabilities of PDA devices. In particular, we experimentally measure and compare the energy consumption and computation time for four different block cipher encryption algorithms – RC2, Blowfish, the eXtended Tiny Encryption Algorithm (XTEA), and the Advanced Encryption Standard (AES) – executing on contemporary PDAs. We measure energy consumption for one device and latency and throughput for three different devices. The experiments consider different transfer sizes from one kilobyte to one megabyte. Based on the results, we develop observations that are intended to increase the awareness of and insight into the costs associated with using encryption algorithms. This paper is divided into five sections. Section 2 discusses related work and briefly describes the encryption algorithms considered in this study. Section 3 describes the experimental setup and evaluation methods. Section 4 discusses the measurements and results. Finally, Section 5 summarizes the work, discusses contributions, and proposes future work.



Download full report
http://googleurl?sa=t&source=web&cd=1&ve...1.127.2393%26rep%3Drep1%26type%3Dpdf&ei=2mdbTuTlOs2srAeK_sS9Cg&usg=AFQjCNGSyJuGiXEwp1CSSS3V5cpERDj59A