Dynamic Merkle Trees for Verifying Privileges in Sensor Networks

Mobile sinks are already used in static sensor networks to facilitate data collection or network management, but it is possible in such networks for mobile sinks to be compromised and mount various insider attacks. Some schemes have been devised recently to limit the privileges of mobile sinks appropriately, but these are secure only when the number of compromised mobile sinks or sensors is below a network-wide threshold. This threshold is determined by a sensor's memory, and is usually too small. We introduce dynamic Merkle trees, and show how to use them to restrict and verify the privilege of mobile sinks. Unlike current schemes, our scheme allows all sensors in a region to simultaneously verify mobile sink privileges, and does not assume a network-wide limit on the number of node compromises. As our security analysis shows, our approach is safe from fabrication attacks, impersonation attacks and replay attacks. Further, our performance analysis shows that we incur very low communication, computation, and storage overheads.