Evaluating Homomorphic Encryption Algorithms for Privacy in Wireless Sensor Networks

The Wireless Sensor Networks (WSNs) consist of the sensor nodes those are severely resource constrained. Hence, the protocols devised for the WSNs employ every possible means of optimizing the associated overhead.The WSNs typically, employ in-network processing that is largely based on the data aggregation operations that aggregate the data into a compact representation for further transmission. However, when aggregating a number of different data packets, the integrity of the data values used for aggregation as well as the security of the aggregator node itself, assume prime importance. Even otherwise, due to the ubiquitous and pervasive deployment of the sensor nodes on one hand and due to the resource intensive nature of security algorithms, it is challenging to devise the security protocols for WSNs. Consequently, the protocols for ensuring secure data aggregation also have to be devised and investigated carefully. One of the approaches to ensure secure data aggregation is to use encrypted sensor data for processing, using homomorphic encryption. However, there are several homomorphic encryption algorithms for conventional networks and the feasibility of applying them in resource constrained environment remains an issue to be investigated. In this paper, we attempt at evaluating the performance of various privacy homomorphism algorithms in WSNs. Our work is motivated by identifying a homomorphic encryption algorithm for WSNs that offers security at the minimum overhead. To the best of our knowledge, ours is the first experimental evaluation of the state-of-the-art privacy homomorphic algorithms for WSNs.