On the practicality of elliptic curve cryptography for medical sensor networks

Biomedical sensors have long been used to monitor patients in hospitals and emergency settings. Until recently, these sensors were connected to their respective monitors through an array of cumbersome wires. Recent research has made it possible for biomedical sensors to communicate over wireless links, creating new opportunities for patients and caregivers. Not only can the hassle of physical wires be eliminated, but wireless communication has enabled the ability for sensors to form ad-hoc mesh networks with each other. With wireless communication, however, effective secure communication must now become a priority in countries like the United States where a patient's privacy is mandated by law. Until now, security in biomedical sensor networks has been provided by symmetric key cryptography. While computationally efficient, pre-shared keys introduce the new obstacles of key management, deployment and protection. Public key cryptography is typically used as a solution to this overhead, but has not been feasible on remote sensors due to their limited computational resources. To this end, TinyECC is a public key algorithm with optimizations for resource constrained hardware platforms. In this paper, we implement a prototype biomedical sensor application implementing TinyECC to secure the wireless communication between sensor nodes, and study the feasilbility of using TinyECC in a real-time sensor network.