Shared Security: How Wireless Sensor Networks Can Benefit from Threshold Cryptography

Wireless sensor networks consist of a huge number of small nodes, communicating wirelessly, to transmit any sort of measured data, like temperature, radiation, etc. At the air interface, unprotected messages can be easily intercepted and modified by an attacker. Traditionally, symmetric cryptography is deployed in sensor networks, due to the nodes being constrained in terms of energy, processing power and memory. If an attacker is now able to extract the secret symmetric key from a single node, the entire (or a huge subset of the) network is compromised. Threshold cryptography is an attractive approach to this problem: by separating the secret into several parts, an attacker has to compromise at least t + 1 nodes to be able to extract a meaningful value. In this work we investigate computational optimizations to the multiparty multiplication protocol of Gennaro, Rabin, and Rabin, thereby improving the running time of certain protocol steps by a factor of up to 6.