Securing Timing Synchronization in Sensor Networks

Time synchronization is critical in sensor networks at many layers of its design. It enables better duty-cycling of the radio, accurate and secure localization, beamforming and other collaborative signal processing. Examples of existing sensor network applications where precise time is needed include: measuring the time-of-flight of sound; distributing an acoustic beam forming array; forming a low-power TDMA radio schedule; integrating a time-series of proximity detections into a velocity estimate; suppressing redundant messages by recognizing duplicate detections of the same event by different sensors; ordered logging of events during system debugging; integrating multi sensor data; or coordinating on future action. Imagine the detrimental affect on the functionality of all these applications, if a malicious adversary is able to abuse the underlying time synchronization protocol. Nodes will have faulty estimates about the location of other nodes. Packets will be lost if the sleep-wakeup schedule of nodes do not intersect. This can further trigger unnecessary packet retransmissions if MAC layer acknowledgements are enabled. It will be trivial for adversaries to perform replay attacks. Collaborative data processing and signal processing techniques will be adversely affected.