Public-key based access control in sensornet

Abstract

Symmetric cryptography has been mostly used in security schemes in sensor networks due to the concern that public key cryptography (PKC) is too expensive for sensor devices. While these schemes are efficient in processing time, they generally require complicated key management, which may introduce high memory and communication overhead. On the contrary, PKC-based schemes have simple and clean key management, but cost more computational time. The recent progress in PKC implementation, specially elliptic curve cryptography (ECC), on sensors motivates us to design a PKC-based security scheme and compare its performance with the symmetric-key counterparts. This paper proposes a practical PKC-based access control for sensor networks, which consists of pairwise key establishment, local access control, and remote access control. We have implemented both cryptographic primitives on commercial off-the-shelf sensor devices. Building the user access control as a case study, we show that PKC-based protocol is more advantageous than those built on symmetric cryptography in terms of the memory usage, message complexity, and security resilience. Meanwhile, our work also provides insights in integrating and designing PKC-based security protocols for sensor networks.