Key Management Building Blocks for Wireless Sensor Networks

Cryptography is the means to ensure data confidentiality, integrity and authentication in wireless sensor networks (WSNs). To use cryptography effectively however, the cryptographic keys need to be managed properly. First of all, the necessary keys need to be distributed to the nodes before the nodes are deployed in the field, in such a way that any two or more nodes that need to communicate securely can establish a session key. Then, the session keys need to be refreshed from time to time to prevent birthday attacks. Finally, in case any of the nodes is found to be compromised, the key ring of the compromised node needs to be revoked and some or all of the compromised keys might need to be replaced. These processes, together with the policies and techniques needed to support them, are called key management. The facts that WSNs (1) are generally not tamper-resistant; (2) operate unattended; (3) communicate in an open medium; (4) have no fixed infrastructure and pre-configured topology; (5) have severe hardware and resource constraints, present unique challenges to key management. In this article, we explore techniques for meeting these challenges. What distinguishes our approach from a routine literature survey is that, instead of comparing various known schemes, we set out to identify the basic cryptographic principles, or building blocks that will allow practitioners to set up their own key management framework using these building blocks.