Confidentiality Protection Schemes for Data Aggregation in Sensor Networks

Abstract

Efficiency and security are two basic requirements for sensor networkdesign. However, these requirements could be sharply conflicting in somescenarios. For example, in-network data aggregation can significantlyreduce communication overhead and thus has been adopted widely as a meansto improve network efficiency; on the other hand, the adoption ofin-network data aggregation prevents the data from being encrypted sincedata being accessible during forwarding is a prerequisite for aggregation.In this paper, we investigate this dilemma by proposing a new series ofsecret perturbation-based schemes that can protect sensor dataconfidentiality without disrupting data aggregation. Extensive analysis andsimulations are also conducted to evaluate the proposed protocols and tostudy the tradeoff between data confidentiality and communicationefficiency.