Improving key predistribution with deployment knowledge in static sensor networks

Abstract

Pairwise key establishment is a fundamental security service for sensor networks. However, establishing pairwise keys in sensor networks is a challenging problem, particularly due to the resource constraints on sensor nodes and the threat of node compromises. This article proposes to use both predeployment and postdeployment knowledge to improve pairwise key predistribution in static sensor networks. By exploiting the predeployment knowledge, this article first develops two key predistribution schemes, a closest pairwise keys scheme and a closest polynomials scheme . The analysis shows that these schemes can achieve better performance if the expected location information is available and that the smaller the deployment error is, the better performance they can achieve. The article then investigates how to use postdeployment knowledge to improve pairwise key predistribution in static sensor networks. The idea is to load an excessive amount of predistributed keys on sensor nodes, prioritize these keys based on sensors' actual locations discovered after deployment, and discard low-priority keys to thwart node compromise attacks. This approach is then used to improve the random subset assignment scheme proposed recently to demonstrate its practicality and effectiveness. The analysis indicates that the postdeployment knowledge can also greatly improve the performance and security of key predistribution.