Energy efficiency in secure wireless sensor networks

Abstract

Cryptography plays an important role in protecting the data security of wireless sensor networks (WSNs). However, it is greatly constrained by the limited energy supply of a sensor node. In this thesis, we focus on the energy efficiency of secure communication in wireless sensor networks (WSNs). Our research considers both the cryptographic algorithms and the cryptographic schemes when link layer security is applied to a WSN. By implementing symmetric key ciphers in both software and hardware, we evaluate the computational energy cost considering both the algorithm characteristics and channel quality. We further explore different factors that affect the communication energy cost of link layer cryptographic schemes, such as the payload size, the mode of operation, the initialization vector distribution, and the communication channel quality. We evaluate the energy efficiency of different cryptographic schemes for data transmission by developing an analysis model that is constructed considering various factors affecting both the computational cost and communication cost. Its appropriateness is further verified and supported by simulation results. We also investigate the energy cost of session key establishment in a WSN. In conclusion, we recommend using a lightweight block cipher, byte-wise substitution permutation network (BSPN), along with the ciphertext feedback scheme for wireless sensor networks to achieve security and better energy efficiency. To enhance the data security further in a WSN, we also suggest using a key distribution protocol based on symmetric key cryptography, which can lead to a robust system with acceptable additional energy consumption.