Bootstrapping trust in networked measurement systems with secure sensors

Aggregation of local sensor observations is a widely used and powerful approach to estimating the global state of large networked systems. However, guaranteeing the integrity of the aggregate remains an open problem, in particular for dynamic networked systems. We consider this problem in a single aggregator model - a system of several networked sensors and one inherently trusted collector - and propose a simple, yet effective, system design for ensuring integrity, supported by a proof-of-concept prototype.